via:
Sunday, October 11, 2026
Saturday, October 10, 2026
Tomorrow is Waiting
If you want the truth, it happened because Anji was feeling lazy. Her AI class wasn't all that interesting, nor was it a field she wanted a career in, so there wasn't any reason she could see for trying especially hard. So she came up with a project that didn't look like too much work, and she picked what looked like the easiest way of doing it. Things just got a little out of hand, after that.
Anji's AI class was taught by a grad student who seemed as bored as her students. It was a graduation requirement for programmers, even though everyone knew AIs, as a field, weren't going anywhere much. In seventy years of computing nobody yet had designed an AI that passed the Turing test, let alone did anything really interesting. No matter the computing power behind them, AIs just couldn't be as complex as a human brain; everyone knew that. Anji and her classmates still needed to know how to use the little crippleware bots that ran traffic lights and production lines, though, and that meant knowing the basics of AI programming. At least well enough to pass the final.
So Anji decided to pick the easiest-looking project off the list of options: Design an AI that mimics the behavior of a public domain character. There was a list of characters to choose from, mostly stuff she'd never heard of. She picked Kermit the Frog because, she figured, there was a ton of footage of Kermit, even if it was mostly fifty years old, and she could just feed old TV shows to a bot until it started acting enough like Kermit to get her a passing grade.
Only it wasn't that easy. For one thing, the bot was too stupid to understand that it was meant to be Kermit. Anji used off-the-shelf open-source language- and image-parsing software, so the bot would understand what it what watching, but she had to write a program to key the bot to Kermit in particular. It took forever. It was actually a pretty good challenge, writing a program to convince the bot that it was Kermit the Frog, that the little fuzzy green thing in the old video was itself—that it had a self, for that matter. She ended up using concepts and bits of code from the other classes she was taking, pulling a few all-nighters at the library with books on AI design, and just plain making stuff up in a few places. Her code wasn't anything like elegant, but Anji found herself liking the project a lot more than she'd expected to, even as it got harder.
She also found herself liking Kermit a lot more than she'd expected to. Anji had never really watched the Muppets before; her parents, like most parents she knew, had treated TV as only slightly less corrupting an influence than refined sugar and gendered toys. But The Muppet Show was really funny—strange, and kind of hokey, but charming all the same. She ended up watching way more of it than she needed just for the project.
Then her friend Brian, who was really into robotics, got wind of what she was doing, and demanded the chance to participate. Apparently he had weird, nostalgic parents who'd actually allowed him to watch TV as a kid, and what he'd mostly watched was Sesame Street and the Muppets, so the chance to make a real live AI-powered Kermitbot was too good to pass up.
Of course, that made more work for Anji. She had finally gotten the bot keyed to Kermit properly, so it didn't get confused every time there was another Muppet on screen that looked vaguely froggy or was voiced by Jim Henson, and it was sucking down footage at a pretty good clip—luckily there was so much to feed it, on top of the movies: hours and hours of TV specials and commercials and interviews and even outtakes, all of it in character. But now she had to write a whole new suite of programs so the little AI could operate a robot body. Anji started to worry about finishing the project by the due date. For that matter, she was getting behind in her other classes, and it would be downright embarrassing to do poorly in them because AI design, of all things, was taking up her time.
The thing was, her little AI was getting kind of interesting. It had started writing its own code about the time she'd gotten it keyed to Kermit properly, which was one of the project requirements, but Anji hadn't expected much more than a few badly parsed lines. Nobody else in her class was getting more than that, but Anji's AI was producing more code all the time. And weird code, too. Anji couldn't really make sense of it, but it was working, apparently: the bot hadn't frozen up or crashed, and it wasn't having any trouble parsing the footage Anji fed it.
Brian finished his robot a couple of days after Anji got through the last of the footage. He presented it to her proudly, like a cat gives you something really good it's killed and expects your praise for it. "Good, isn't he?" Brian asked, beaming at her, and Anji had to admit it was convincing. Brian had really gone all out: the little robot was fully articulated ("Enough to play the banjo!" Brian pointed out), and perfectly accurate, with plenty of internal memory built in, and a wireless charger. It didn't even need to be plugged in to upload Anji's code. Not that most of it was really Anji's, anymore. She was starting to wonder if this project wasn't getting away from her a little.
The one change Brian had made, in designing his robot, was to give it eyelids. He said it was creepy without them. So when Anji hit the key that uploaded her code, the first sign she had that it had worked was when Kermit gave a couple of slow, sleepy blinks. "Oh," he said, sitting up (Anji was glad to see she'd done a good job with the movement programs), "hello there."
"Hi, Kermit!" Brian said, all dorkily excited. "I'm Brian. It's really nice to meet you."
He elbowed Anji. "Uh, hi," she said. "I'm Anjali. Anji, really."
"Hello, Anji," Kermit said. "Pleased to meet you. I'm Kermit the Frog," and hey, that sounded exactly right. Anji was totally getting an A.
Anji let Brian keep talking to Kermit, and went to check her computer to make sure everything had uploaded okay. It looked fine: everything running smooth. Only the bot was still writing new code, even as it chatted with Brian. Huh. Anji looked back over at them; Kermit had said something that was making Brian laugh really, really hard. Bots weren't supposed to be very good at telling jokes, were they? They'd covered that in class: how AIs never really seemed to get how jokes worked, and even AIs designed to tell them mostly just produced a sort of unfunny word salad. Maybe Kermit was just quoting the jokes from the footage she'd fed him. AIs could mimic like that, although if she'd built a bot that could mimic good comic timing she deserved more than just an A.
In the weeks that followed, it got harder to treat Kermit like a school project. He spent a lot of his time with Brian, who claimed to need to do a bunch of unspecified adjustments to the robot, although this mostly seemed to entail Kermit being shown off to all Brian's friends. Anji didn't mind it too much, though, because it gave her more time to try and puzzle out Kermit's code, and also it meant that Kermit acquired a very small banjo and several sets of little clothes from Muppet fans among Brian's friends. And that seemed to make Kermit happy.
That was the freaky thing: Anji had designed a bot that could seem to be happy. She wasn't supposed to be able to do that. She was way, way outside the parameters of her project now, into territory that people who studied AI for a living hadn't covered anywhere Anji could find. Because Kermit could, in fact, make jokes—and if he was mimicking them, the originals weren't in the footage Anji had fed him—and he could noodle around on the banjo in a way that sounded nothing like the precision of music-playing AIs Anji had heard. And he could also do things that freaked Anji out on a deep and meaningful personal level, like the afternoon when Kermit, perched on the edge of the bed in Anji's dorm, stopped strumming his banjo and sighed wistfully.
"You know, I sure do miss Fozzie," he announced, and Anji stopped typing mid-keystroke.
"What did you say?" Anji asked, trying not to sound as startled as she felt.
"Oh, it's not that I don't like it here, Anji. You and Brian are awfully nice. But Fozzie's my best friend, you know? After a while, you get to miss things. The squeak of a rubber chicken. The smell of custard pie on fur. Little things like that."
He sighed again, and went back to strumming his banjo. Anji waited five minutes, excused herself, and ran full-tilt across campus to Brian's dorm.
He answered the door, looking concerned. Well, Anji had been hammering on it pretty hard. "What's the matter? Is Kermit okay?"
"Brian, I think we invented sentient AI." Anji tried not to sound like she was panicking. She totally was, though. "We weren't supposed to invent sentient AI! I was just supposed to get a passing grade! Now there's an artificial life-form in my dorm room who plays the banjo!"
"Whoa. Calm down. Why are you freaking out now? Kermit hasn't gotten any more sentient than he was last week, has he? And why is it such a big deal if he is?"
"People have been trying to build a sentient AI for like seventy years, Brian. And I knocked one together out of spare parts for a freshman project in a class I didn't even want to take!" Anji wasn't sure how people were going to react, but she didn't think it would be good. The grad student who taught her class would probably be pissed. "No one's going to believe I actually programmed him, or that he's really sentient. But he just told me he misses his friend and made a couple of novel jokes that made sense, so I'm pretty sure I've created life. And I bet I'm going to get in trouble for it."
Brian, damn him, thought she was overreacting. Worse, he thought she was mostly worried about her grade. They ended up fighting over it, getting into a yelling match that drew Brian's RA in to tell them they were damaging the rest of the floor's calm. Anji really didn't like Brian's RA.
Anji trudged back across campus to her dorm that night in a gloomy frame of mind. Sure, it was pretty cool that she had created sentient AI, but she was afraid it would cause more problems than she really knew how to handle. There was the issue of convincing people Kermit was really sentient, just for starters, and then what was he supposed to do with himself, if people ever believed he was for real? He was just a little frog in a big world, when you got down to it.
Lost in her own thoughts, Anji didn't hear the music until she was nearly back to her dorm. When the sound finally made its way through her thick skull, she paused outside her door, and just listened for a minute. Kermit was singing a song.
It wasn't anything Anji had heard before. The lyrics were sweet and simple, all about looking towards the future, and how it was always just a day away. "I won't miss yesterday," Kermit sang, "because I can see—tomorrow is waiting for me." He strummed a few more chords on the banjo, and fell silent.
Anji pushed the door open. "I liked your song, Kermit," she said.
"Thanks, Anji," Kermit said. "It just kinda came to me, you know? That's why I like singing."
"Yeah," Anji said. She thought about her project deadline, three days away, and the other homework she wasn't getting done. Then she sat down at her desk and called up a fresh copy of the generic AI, the same blank template she'd started from with Kermit, and got to work keying it to Fozzie.
She wasn't anything like done, three days later, when it was time to present her project, but she'd gotten a lot of good practice with Kermit, and she thought she could have Fozzie up and running inside of two weeks. Kermit walked with her to class, carrying his banjo slung across his back, and Anji ignored the funny looks they got from the other students passing them. She was busy with a sudden, unexpected flurry of guilt: what right, she thought, did she have to show Kermit off to her class like—like some kind of show frog? If he was sentient, he deserved better. Just because he didn't seem to mind—was, in fact, excited to be performing for an audience—didn't mean that Anji was doing the right thing.
But right or wrong, if she didn't show up with something to show for a semester's worth of work, her GPA would be toast. Anji felt guilty, but that didn't stop her from being practical. And she could hope for allies among her classmates, maybe. Once they saw Kermit, they might understand.
Or she could get in a lot of trouble. That was the thought at the top of her mind as Anji came into the classroom, and nervously eyed Malika, the grad student waiting at the front of the room. A few other students had already arrived, most of them carrying the tablets or laptops they'd demonstrate their own projects with. A few had robots, but theirs were little bug-like creatures or wheeled rovers.
To her surprise, Malika brightened as soon as she saw Kermit, and came over to talk to Anji. "You did Kermit?" she said, sounding delighted. "Wow, he looks really great. Just like the real thing. Who built him?"
"Um," Anji said, already embarrassed to be talking like Kermit wasn't there. "There's something I'd like to talk to you about, actually. In private?"
"Sure, sure, after class," Malika said. "I can't wait to see your presentation!"
Somehow, it was worse than if Malika hadn't been interested at all. Kermit looked up at her, concern showing on his small green face. "Are you all right, Anji?"
She hadn't known how to talk to Kermit about the problem. And now it seemed like it was too late. "Just nervous, that's all," she lied.
"You don't have to be nervous," Kermit said. "I mean, sharing something with an audience for the first time is always a little scary, but I've got lots of practice. You don't need to worry about me."
Yes, I do, Anji thought, but she didn't say it.
Kermit didn't seem to be bothered by the fact that the other presentations were all about code and AIs and made frequent mention of bots and programming. Anji wished she'd gotten up the nerve to talk to him about what, exactly, he thought he was—if he knew he was a robot, if he understood he was a sentient AI, if he even got what any of that meant. But she'd been too scared to do it, and Brian had been too excited about his new little green friend. She felt miserably like she'd betrayed Kermit's trust.
When Kermit's—when Anji's turn came, Kermit strolled down to the front of the room and hopped easily onto Malika's desk, settling his banjo on his lap. "Hi-ho everyone," he began. "Kermit the Frog here. My friend Anji asked me to put on a show for you. I haven't got the backup I usually have—and anyway, I don't think there's room in here for a chicken chorus, or a penguin orchestra, or a cannon—but I thought I might sing you a song. I hope you all like it."
The class, who had giggled a little at Kermit's joke, fell quiet as he began to sing. It was the same song Anji had heard him working on before, but he'd changed some of the lyrics, and the arrangement wasn't quite the same. The new version was a little better, actually, to Anji's ear. She looked anxiously at her classmates' faces, at Malika, as Kermit sang the chorus again. He played a little flourish on his banjo, sang "Tomorrow is waiting for me" one last time, and strummed a final chord.
There was silence in the classroom, for a long moment. Then someone started clapping, and the rest of the class joined in, and Anji smiled with relief until she saw that Malika wasn't clapping. She looked serious, and thoughtful.
After class, in the empty lecture hall, Malika still looked grave. "Anji, you've put me in kind of a difficult position," she said. "You're obviously a talented programmer, but the project requirements were pretty clear. You weren't supposed to program a performance, you were supposed to get some novel behavior out of your AI."
"Um," said Anji. "This is what I wanted to talk to you about before class, actually. I was kind of afraid of this. See, I didn't program that."
"Then who did?"
"No one did! Kermit came up with it on his own. I'm tone-deaf, anyway; I can't write music."
"Aw, I wouldn't say tone-deaf, Anji," Kermit said. "I've heard you humming along a few times. Tone-confused, maybe, but I bet with a little practice you could get better."
Malika stared at Kermit. Then she said, "Anji, can I have a minute alone with your—with Kermit?"
Anji looked anxiously down at him. "She just wants to ask you a few questions, I think," she said. "Is that okay?"
"No problem," said Kermit. "I interview well."
Anji sat with her back to the wall outside the classroom, the minutes stretching out like taffy. She watched the other students passing by, and wondered if any of them had ever managed to get themselves into a fix like this.
Well, no. Probably not. None of them had invented sentient AI, after all. That was pretty much a one-time thing, unless she got Fozzie off the ground.
Finally, the door opened, and Malika leaned out into the hall. She looked puzzled, like she'd just eaten something and wasn't sure yet if she liked the taste. "Okay, either you've spent the last three months doing nothing but program in responses to every conceivable question, or he's as close to sentient as any AI I've seen. Either way, you must have been seriously slacking off at the beginning of the semester, because your early assignments don't reflect this level of dedication. How the hell did you do it?"
"It was an accident!" Anji said weakly. "I just kept feeding him footage of Kermit, until he kind of was Kermit. I can show you my notes?"
"I think you'd better," Malika said, and stood aside so Anji could come back into the lecture hall.
The next few weeks were confusing, but in a good way. Anji had a truly terrifying meeting with her AI professor, which was mitigated a little by Malika going to bat for her. Her professor didn't come around as easily—apparently he wasn't a Muppet fan—but Anji's code made his eyebrows go up in a promising way, and when they left his office he was already emailing some other AI experts.
Meanwhile, Kermit was becoming something of a star on campus. People were always excited to meet him—some because they were meeting what might be a sentient AI, and other just because they were meeting Kermit the Frog. Brian started working on the bot for Fozzie, and Anji hit the stage where her Fozzie program started writing its own code. This time it was a lot better-documented, with Malika practically peering over her shoulder as the first lines appeared.
Some friends of Brian's at another school got in touch with her, asking if they could use her keying programs to bring Gonzo to life; they apparently had a build team ready to go. Anji said yes. That set off a wave of Internet chatter. Up to now, there hadn't really been any media attention—her professor wanted to wait until enough experts had met with Kermit and agreed that he was sentient. Or "close enough to sentient to fool me," which was his begrudging way of putting it.
But Brian's friends were blogging the whole build process, and the attention they drew eventually found its way back to Anji. For the first time, she found herself fielding interview requests, and a local news team actually came out in person to film her and Kermit talking to their reporter.
Kermit took the whole thing in stride. Well, he would be used to media attention, Anji figured; he had lots of experience dealing with famous people and reporters. When they weren't being interviewed, Kermit spent his time playing music, writing in a little notebook, walking around campus with Anji and talking to people, hanging out with Brian and his friends. He seemed happy, especially when Anji told him Fozzie would be joining them soon. But part of Anji wasn't convinced that everything was okay.
Finally, she got up the nerve to talk about it. "Kermit," she asked, "what—exactly what are you?"
Anji's AI class was taught by a grad student who seemed as bored as her students. It was a graduation requirement for programmers, even though everyone knew AIs, as a field, weren't going anywhere much. In seventy years of computing nobody yet had designed an AI that passed the Turing test, let alone did anything really interesting. No matter the computing power behind them, AIs just couldn't be as complex as a human brain; everyone knew that. Anji and her classmates still needed to know how to use the little crippleware bots that ran traffic lights and production lines, though, and that meant knowing the basics of AI programming. At least well enough to pass the final.
Only it wasn't that easy. For one thing, the bot was too stupid to understand that it was meant to be Kermit. Anji used off-the-shelf open-source language- and image-parsing software, so the bot would understand what it what watching, but she had to write a program to key the bot to Kermit in particular. It took forever. It was actually a pretty good challenge, writing a program to convince the bot that it was Kermit the Frog, that the little fuzzy green thing in the old video was itself—that it had a self, for that matter. She ended up using concepts and bits of code from the other classes she was taking, pulling a few all-nighters at the library with books on AI design, and just plain making stuff up in a few places. Her code wasn't anything like elegant, but Anji found herself liking the project a lot more than she'd expected to, even as it got harder.
She also found herself liking Kermit a lot more than she'd expected to. Anji had never really watched the Muppets before; her parents, like most parents she knew, had treated TV as only slightly less corrupting an influence than refined sugar and gendered toys. But The Muppet Show was really funny—strange, and kind of hokey, but charming all the same. She ended up watching way more of it than she needed just for the project.
Then her friend Brian, who was really into robotics, got wind of what she was doing, and demanded the chance to participate. Apparently he had weird, nostalgic parents who'd actually allowed him to watch TV as a kid, and what he'd mostly watched was Sesame Street and the Muppets, so the chance to make a real live AI-powered Kermitbot was too good to pass up.
Of course, that made more work for Anji. She had finally gotten the bot keyed to Kermit properly, so it didn't get confused every time there was another Muppet on screen that looked vaguely froggy or was voiced by Jim Henson, and it was sucking down footage at a pretty good clip—luckily there was so much to feed it, on top of the movies: hours and hours of TV specials and commercials and interviews and even outtakes, all of it in character. But now she had to write a whole new suite of programs so the little AI could operate a robot body. Anji started to worry about finishing the project by the due date. For that matter, she was getting behind in her other classes, and it would be downright embarrassing to do poorly in them because AI design, of all things, was taking up her time.
The thing was, her little AI was getting kind of interesting. It had started writing its own code about the time she'd gotten it keyed to Kermit properly, which was one of the project requirements, but Anji hadn't expected much more than a few badly parsed lines. Nobody else in her class was getting more than that, but Anji's AI was producing more code all the time. And weird code, too. Anji couldn't really make sense of it, but it was working, apparently: the bot hadn't frozen up or crashed, and it wasn't having any trouble parsing the footage Anji fed it.
Brian finished his robot a couple of days after Anji got through the last of the footage. He presented it to her proudly, like a cat gives you something really good it's killed and expects your praise for it. "Good, isn't he?" Brian asked, beaming at her, and Anji had to admit it was convincing. Brian had really gone all out: the little robot was fully articulated ("Enough to play the banjo!" Brian pointed out), and perfectly accurate, with plenty of internal memory built in, and a wireless charger. It didn't even need to be plugged in to upload Anji's code. Not that most of it was really Anji's, anymore. She was starting to wonder if this project wasn't getting away from her a little.
The one change Brian had made, in designing his robot, was to give it eyelids. He said it was creepy without them. So when Anji hit the key that uploaded her code, the first sign she had that it had worked was when Kermit gave a couple of slow, sleepy blinks. "Oh," he said, sitting up (Anji was glad to see she'd done a good job with the movement programs), "hello there."
"Hi, Kermit!" Brian said, all dorkily excited. "I'm Brian. It's really nice to meet you."
He elbowed Anji. "Uh, hi," she said. "I'm Anjali. Anji, really."
"Hello, Anji," Kermit said. "Pleased to meet you. I'm Kermit the Frog," and hey, that sounded exactly right. Anji was totally getting an A.
Anji let Brian keep talking to Kermit, and went to check her computer to make sure everything had uploaded okay. It looked fine: everything running smooth. Only the bot was still writing new code, even as it chatted with Brian. Huh. Anji looked back over at them; Kermit had said something that was making Brian laugh really, really hard. Bots weren't supposed to be very good at telling jokes, were they? They'd covered that in class: how AIs never really seemed to get how jokes worked, and even AIs designed to tell them mostly just produced a sort of unfunny word salad. Maybe Kermit was just quoting the jokes from the footage she'd fed him. AIs could mimic like that, although if she'd built a bot that could mimic good comic timing she deserved more than just an A.
In the weeks that followed, it got harder to treat Kermit like a school project. He spent a lot of his time with Brian, who claimed to need to do a bunch of unspecified adjustments to the robot, although this mostly seemed to entail Kermit being shown off to all Brian's friends. Anji didn't mind it too much, though, because it gave her more time to try and puzzle out Kermit's code, and also it meant that Kermit acquired a very small banjo and several sets of little clothes from Muppet fans among Brian's friends. And that seemed to make Kermit happy.
That was the freaky thing: Anji had designed a bot that could seem to be happy. She wasn't supposed to be able to do that. She was way, way outside the parameters of her project now, into territory that people who studied AI for a living hadn't covered anywhere Anji could find. Because Kermit could, in fact, make jokes—and if he was mimicking them, the originals weren't in the footage Anji had fed him—and he could noodle around on the banjo in a way that sounded nothing like the precision of music-playing AIs Anji had heard. And he could also do things that freaked Anji out on a deep and meaningful personal level, like the afternoon when Kermit, perched on the edge of the bed in Anji's dorm, stopped strumming his banjo and sighed wistfully.
"You know, I sure do miss Fozzie," he announced, and Anji stopped typing mid-keystroke.
"What did you say?" Anji asked, trying not to sound as startled as she felt.
"Oh, it's not that I don't like it here, Anji. You and Brian are awfully nice. But Fozzie's my best friend, you know? After a while, you get to miss things. The squeak of a rubber chicken. The smell of custard pie on fur. Little things like that."
He sighed again, and went back to strumming his banjo. Anji waited five minutes, excused herself, and ran full-tilt across campus to Brian's dorm.
He answered the door, looking concerned. Well, Anji had been hammering on it pretty hard. "What's the matter? Is Kermit okay?"
"Brian, I think we invented sentient AI." Anji tried not to sound like she was panicking. She totally was, though. "We weren't supposed to invent sentient AI! I was just supposed to get a passing grade! Now there's an artificial life-form in my dorm room who plays the banjo!"
"Whoa. Calm down. Why are you freaking out now? Kermit hasn't gotten any more sentient than he was last week, has he? And why is it such a big deal if he is?"
"People have been trying to build a sentient AI for like seventy years, Brian. And I knocked one together out of spare parts for a freshman project in a class I didn't even want to take!" Anji wasn't sure how people were going to react, but she didn't think it would be good. The grad student who taught her class would probably be pissed. "No one's going to believe I actually programmed him, or that he's really sentient. But he just told me he misses his friend and made a couple of novel jokes that made sense, so I'm pretty sure I've created life. And I bet I'm going to get in trouble for it."
Brian, damn him, thought she was overreacting. Worse, he thought she was mostly worried about her grade. They ended up fighting over it, getting into a yelling match that drew Brian's RA in to tell them they were damaging the rest of the floor's calm. Anji really didn't like Brian's RA.
Anji trudged back across campus to her dorm that night in a gloomy frame of mind. Sure, it was pretty cool that she had created sentient AI, but she was afraid it would cause more problems than she really knew how to handle. There was the issue of convincing people Kermit was really sentient, just for starters, and then what was he supposed to do with himself, if people ever believed he was for real? He was just a little frog in a big world, when you got down to it.
Lost in her own thoughts, Anji didn't hear the music until she was nearly back to her dorm. When the sound finally made its way through her thick skull, she paused outside her door, and just listened for a minute. Kermit was singing a song.
It wasn't anything Anji had heard before. The lyrics were sweet and simple, all about looking towards the future, and how it was always just a day away. "I won't miss yesterday," Kermit sang, "because I can see—tomorrow is waiting for me." He strummed a few more chords on the banjo, and fell silent.
Anji pushed the door open. "I liked your song, Kermit," she said.
"Thanks, Anji," Kermit said. "It just kinda came to me, you know? That's why I like singing."
"Yeah," Anji said. She thought about her project deadline, three days away, and the other homework she wasn't getting done. Then she sat down at her desk and called up a fresh copy of the generic AI, the same blank template she'd started from with Kermit, and got to work keying it to Fozzie.
She wasn't anything like done, three days later, when it was time to present her project, but she'd gotten a lot of good practice with Kermit, and she thought she could have Fozzie up and running inside of two weeks. Kermit walked with her to class, carrying his banjo slung across his back, and Anji ignored the funny looks they got from the other students passing them. She was busy with a sudden, unexpected flurry of guilt: what right, she thought, did she have to show Kermit off to her class like—like some kind of show frog? If he was sentient, he deserved better. Just because he didn't seem to mind—was, in fact, excited to be performing for an audience—didn't mean that Anji was doing the right thing.
But right or wrong, if she didn't show up with something to show for a semester's worth of work, her GPA would be toast. Anji felt guilty, but that didn't stop her from being practical. And she could hope for allies among her classmates, maybe. Once they saw Kermit, they might understand.
Or she could get in a lot of trouble. That was the thought at the top of her mind as Anji came into the classroom, and nervously eyed Malika, the grad student waiting at the front of the room. A few other students had already arrived, most of them carrying the tablets or laptops they'd demonstrate their own projects with. A few had robots, but theirs were little bug-like creatures or wheeled rovers.
To her surprise, Malika brightened as soon as she saw Kermit, and came over to talk to Anji. "You did Kermit?" she said, sounding delighted. "Wow, he looks really great. Just like the real thing. Who built him?"
"Um," Anji said, already embarrassed to be talking like Kermit wasn't there. "There's something I'd like to talk to you about, actually. In private?"
"Sure, sure, after class," Malika said. "I can't wait to see your presentation!"
Somehow, it was worse than if Malika hadn't been interested at all. Kermit looked up at her, concern showing on his small green face. "Are you all right, Anji?"
She hadn't known how to talk to Kermit about the problem. And now it seemed like it was too late. "Just nervous, that's all," she lied.
"You don't have to be nervous," Kermit said. "I mean, sharing something with an audience for the first time is always a little scary, but I've got lots of practice. You don't need to worry about me."
Yes, I do, Anji thought, but she didn't say it.
Kermit didn't seem to be bothered by the fact that the other presentations were all about code and AIs and made frequent mention of bots and programming. Anji wished she'd gotten up the nerve to talk to him about what, exactly, he thought he was—if he knew he was a robot, if he understood he was a sentient AI, if he even got what any of that meant. But she'd been too scared to do it, and Brian had been too excited about his new little green friend. She felt miserably like she'd betrayed Kermit's trust.
When Kermit's—when Anji's turn came, Kermit strolled down to the front of the room and hopped easily onto Malika's desk, settling his banjo on his lap. "Hi-ho everyone," he began. "Kermit the Frog here. My friend Anji asked me to put on a show for you. I haven't got the backup I usually have—and anyway, I don't think there's room in here for a chicken chorus, or a penguin orchestra, or a cannon—but I thought I might sing you a song. I hope you all like it."
The class, who had giggled a little at Kermit's joke, fell quiet as he began to sing. It was the same song Anji had heard him working on before, but he'd changed some of the lyrics, and the arrangement wasn't quite the same. The new version was a little better, actually, to Anji's ear. She looked anxiously at her classmates' faces, at Malika, as Kermit sang the chorus again. He played a little flourish on his banjo, sang "Tomorrow is waiting for me" one last time, and strummed a final chord.
There was silence in the classroom, for a long moment. Then someone started clapping, and the rest of the class joined in, and Anji smiled with relief until she saw that Malika wasn't clapping. She looked serious, and thoughtful.
After class, in the empty lecture hall, Malika still looked grave. "Anji, you've put me in kind of a difficult position," she said. "You're obviously a talented programmer, but the project requirements were pretty clear. You weren't supposed to program a performance, you were supposed to get some novel behavior out of your AI."
"Um," said Anji. "This is what I wanted to talk to you about before class, actually. I was kind of afraid of this. See, I didn't program that."
"Then who did?"
"No one did! Kermit came up with it on his own. I'm tone-deaf, anyway; I can't write music."
"Aw, I wouldn't say tone-deaf, Anji," Kermit said. "I've heard you humming along a few times. Tone-confused, maybe, but I bet with a little practice you could get better."
Malika stared at Kermit. Then she said, "Anji, can I have a minute alone with your—with Kermit?"
Anji looked anxiously down at him. "She just wants to ask you a few questions, I think," she said. "Is that okay?"
"No problem," said Kermit. "I interview well."
Anji sat with her back to the wall outside the classroom, the minutes stretching out like taffy. She watched the other students passing by, and wondered if any of them had ever managed to get themselves into a fix like this.
Well, no. Probably not. None of them had invented sentient AI, after all. That was pretty much a one-time thing, unless she got Fozzie off the ground.
Finally, the door opened, and Malika leaned out into the hall. She looked puzzled, like she'd just eaten something and wasn't sure yet if she liked the taste. "Okay, either you've spent the last three months doing nothing but program in responses to every conceivable question, or he's as close to sentient as any AI I've seen. Either way, you must have been seriously slacking off at the beginning of the semester, because your early assignments don't reflect this level of dedication. How the hell did you do it?"
"It was an accident!" Anji said weakly. "I just kept feeding him footage of Kermit, until he kind of was Kermit. I can show you my notes?"
"I think you'd better," Malika said, and stood aside so Anji could come back into the lecture hall.
The next few weeks were confusing, but in a good way. Anji had a truly terrifying meeting with her AI professor, which was mitigated a little by Malika going to bat for her. Her professor didn't come around as easily—apparently he wasn't a Muppet fan—but Anji's code made his eyebrows go up in a promising way, and when they left his office he was already emailing some other AI experts.
Meanwhile, Kermit was becoming something of a star on campus. People were always excited to meet him—some because they were meeting what might be a sentient AI, and other just because they were meeting Kermit the Frog. Brian started working on the bot for Fozzie, and Anji hit the stage where her Fozzie program started writing its own code. This time it was a lot better-documented, with Malika practically peering over her shoulder as the first lines appeared.
Some friends of Brian's at another school got in touch with her, asking if they could use her keying programs to bring Gonzo to life; they apparently had a build team ready to go. Anji said yes. That set off a wave of Internet chatter. Up to now, there hadn't really been any media attention—her professor wanted to wait until enough experts had met with Kermit and agreed that he was sentient. Or "close enough to sentient to fool me," which was his begrudging way of putting it.
But Brian's friends were blogging the whole build process, and the attention they drew eventually found its way back to Anji. For the first time, she found herself fielding interview requests, and a local news team actually came out in person to film her and Kermit talking to their reporter.
Kermit took the whole thing in stride. Well, he would be used to media attention, Anji figured; he had lots of experience dealing with famous people and reporters. When they weren't being interviewed, Kermit spent his time playing music, writing in a little notebook, walking around campus with Anji and talking to people, hanging out with Brian and his friends. He seemed happy, especially when Anji told him Fozzie would be joining them soon. But part of Anji wasn't convinced that everything was okay.
Finally, she got up the nerve to talk about it. "Kermit," she asked, "what—exactly what are you?"
by Holly Mintzer, Strange Horizons | Read more:
Image: The Muppets
[ed. Repost (10/30/2023). Cute story. As relevant as ever.]
Agricultural ‘Guest Workers’ in WA Describe Their Lives and Ambitions
The majority of the 400,000 or so agricultural “guest workers” in the United States last year were Mexican men who left the border on chartered buses and rolled, almost nonstop, to American farms and orchards. The first worker I met — in Mattawa, Grant County, — was young and, according to Wendy Lopez, who manages the housing complex where he’d been dropped off a few hours earlier, here on his own. This was supposed to be his first big money-making day in America, but so far, things weren’t going well for him.
Wendy got on her phone to make an appointment for him at a mental health clinic. Away from home for the first time, she told me, he was overwhelmed by anxious thoughts and dark emotions. For now he stood in a hallway, looking at the floor.
Mattawa is in sage steppe desert country. The guest worker housing complex Wendy Lopez manages — built and operated by the Washington Growers League, a nonprofit association — was designed so that contingents of men can live in dormlike units, sharing sleeping quarters, bathrooms, elbow room and kitchens. Wendy led me into one that was empty due to disrepair, turned on the air conditioning and the lights in that order, then went out and soon came back with two men who’d agreed to be interviewed: Odilon, 50, and his son Víctor, 25, from the state of San Luis Potosi in Central Mexico.
We shook hands, then sat. Neither Odilon nor Víctor asked me why I wanted to interview them. I explained myself anyway, thinking I should. “People drive by you every day while you’re working,” I said — or something like that, with Wendy interpreting — “but they don’t know anything about you. Do you think it would be good for them to know some things about you? If you do, tell me what they are and I’ll write them down.”
“H-2A worker” isn’t any clearer than “guest worker,” though, as a term pointing at a reality. “H-2A” is a visa category, like “B-2” for tourists. The United States offers four different “H” visas, all of which allow holders to work here for set durations. There is, for example, an H-1B visa often used for hiring in academia, medicine and technology, among other fields, and an H-2B for landscapers and construction workers. The people I interviewed hold H-2A visas, which allow them to come here for agricultural work. All were hired by labor contractors in Mexico and live here in housing provided by the orchardists who employ them. Some orchardists pay the Growers League or other contractors to provide housing; others provide it on private land.
According to the U.S. Department of Labor, almost 36,000 H-2A positions were certified in Washington state for 2024. Washington accounts for more than 9% of H-2A visas nationally, and Washington employers are among the top 10 in the nation in H-2A hiring.
Odilon is a stocky, broad-faced man whose close cropped black hair stands up like bristles. After we’d spent some time together calculating, I understood that in Mexico he makes about $600 a year doing construction work, and that Víctor makes about $400 doing, as Víctor himself put it while chuckling, a little of everything. Víctor is the oldest of six children, and some of them earn money too, and so does Víctor’s mother. In the end, the family of eight makes about $1,500 a year in Mexico. That didn’t make sense to me at first, because Odilon had said earlier that his family spends $100 a week on food alone, and when I did the math on that it seemed to say they could eat for only 15 weeks every year, assuming they spent all their money on food. I was explicating the financial mystery of this aloud when Víctor cut me off to say, “That’s why we come here.”
Odilon said he and Víctor have been coming to Mattawa as guest workers for the past six years on six-month contracts. He added that he has three daughters legally eligible to be guest workers, but since they were female, he explained, their chances of obtaining visas were slim. Among his family members, only he and Víctor come year after year for half-year stints of earning dollars. Together they bring home $30,000 annually, he said, and who would pass that up? [...]
It didn’t work. Wirtz was only able to recruit 3,300 young athletes. Some quit after a few days; others went on strike. As for the Braceros, a lot of them crossed the border in 1965 and returned to their American jobs. Farm work was done that season mostly by the same people who’d done it before, but now it was illegal for them to be here.
Bad policy often repeats itself. At a news conference last summer in Washington D.C., the federal government’s Secretary of Agriculture, Brooke Rollins, pledged a 100% American workforce. “There are 34 million able-bodied adults in our Medicaid program,” she said. “There are plenty of workers in America.”
Odilon and Víctor support seven people on the $30,000 they jointly make annually as H-2A workers. If the two of them were among Rollins’ able-bodied Americans on Medicaid, their family would be living far below the federal poverty line, which for them would be $47,340 a year.
You’re looking at a “FARMWORKERS WANTED” ad. Fifteen thousand dollars for six months. Yes or no? Take the job or not? Now pretend you’re Odilon and Víctor in Mexico looking at the same ad.
American food security depends on people who cross its southern border. H-2A visas exist so they don’t stay.
We went over the numbers together: in Mexico, Daniel makes about $2,500 working as a lawyer each year. In the U.S., he makes another $8,000 picking fruit. He owns a house, but he and his family moved out after living there for only 15 days because, he said, “someone fired a gun at us.” He rents it out now, so there’s income from that, too.
Meanwhile, there’s broad support in government and agribusiness for the H-2 program. Its growth, supporters argue, would allow for the deportation of workers here illegally while insuring a reliable and stable workforce. This, though, is just their short-term solution. “Ultimately,” Secretary of Agriculture Brooke Rollins said last July, “the answer on this is automation.”
Wendy got on her phone to make an appointment for him at a mental health clinic. Away from home for the first time, she told me, he was overwhelmed by anxious thoughts and dark emotions. For now he stood in a hallway, looking at the floor.
We shook hands, then sat. Neither Odilon nor Víctor asked me why I wanted to interview them. I explained myself anyway, thinking I should. “People drive by you every day while you’re working,” I said — or something like that, with Wendy interpreting — “but they don’t know anything about you. Do you think it would be good for them to know some things about you? If you do, tell me what they are and I’ll write them down.”
****
“Guest worker” is an oxymoron. Guests don’t work; to the contrary, hosts do. “Guest worker” is an odd way to refer to people who earn wages by crossing a border in two directions annually. Most come into the U.S. in early spring and leave in late fall.There was a time — the Great Depression — when the United States didn’t need guest workers. There were enough impoverished people inside our own borders to do all the poorly paid agricultural work that needed doing until, in the early 1940s, millions of them left the country in uniform. At that point the U.S. turned to Mexico to meet its wartime farm labor needs, bringing people across the border on short-term work contracts. For several decades this arrangement went forward as the Bracero Program. Today it’s known as H-2A.
According to the U.S. Department of Labor, almost 36,000 H-2A positions were certified in Washington state for 2024. Washington accounts for more than 9% of H-2A visas nationally, and Washington employers are among the top 10 in the nation in H-2A hiring.
****
Víctor didn’t say much. Every once in a while his father, Odilon, looked at him to see if he agreed about something, or wanted to corroborate, embellish or modify something, but mostly Víctor didn’t. Instead he sat patiently while his father talked.
Odilon said he and Víctor have been coming to Mattawa as guest workers for the past six years on six-month contracts. He added that he has three daughters legally eligible to be guest workers, but since they were female, he explained, their chances of obtaining visas were slim. Among his family members, only he and Víctor come year after year for half-year stints of earning dollars. Together they bring home $30,000 annually, he said, and who would pass that up? [...]
After the Bracero guest worker program was discontinued in 1964 — largely due to pressure from labor unions — Secretary of Labor W. Willard Wirtz tried to replace the 180,000 Mexican workers who would not be coming back for the 1965 farm season with American high school boys through a program he called the A-TEAM, which was an acronym for Athletes in Temporary Employment as Agricultural Manpower.
It didn’t work. Wirtz was only able to recruit 3,300 young athletes. Some quit after a few days; others went on strike. As for the Braceros, a lot of them crossed the border in 1965 and returned to their American jobs. Farm work was done that season mostly by the same people who’d done it before, but now it was illegal for them to be here.
Odilon and Víctor support seven people on the $30,000 they jointly make annually as H-2A workers. If the two of them were among Rollins’ able-bodied Americans on Medicaid, their family would be living far below the federal poverty line, which for them would be $47,340 a year.
You’re looking at a “FARMWORKERS WANTED” ad. Fifteen thousand dollars for six months. Yes or no? Take the job or not? Now pretend you’re Odilon and Víctor in Mexico looking at the same ad.
American food security depends on people who cross its southern border. H-2A visas exist so they don’t stay.
*****
Odilon and Víctor left, and in came Daniel, 31, a slight and gentle man who, while soft-spoken, isn’t reticent. Daniel, in fact, spoke unilaterally often, and extensively too, as if there were things that needed to be understood that I was failing to ask him about. [...]
I told Daniel that Americans would be surprised to hear that a Mexican lawyer comes to the United States every year to pick fruit for six months. It happens, he explained, because good-paying legal jobs in Mexico are hard to come by. Daniel has a degree and some experience but, in his words, “doesn’t know the right people.” He’d been coming to the U.S. to pick fruit for four years, and hoped that, if he did so for a few more, maybe between stints he would impress the right people and land a good paying law job at home.
Daniel said he’d agreed to be interviewed because he wanted to meet someone who had an important job. He wanted to talk about important things, he explained, with important people. What did he mean, I asked, by important people with important jobs? Lawyers, politicians, professors, journalists, people like that, he answered. He wanted to listen to how they spoke and ponder how their minds worked.
Daniel told me I should read his favorite book, “El Hombre Mediocre,” or “The Mediocre Man,” written in 1913 by an Argentine named José Ingenieros. It explained everything, he said ambiguously. Later I found a bad English translation. It’s philosophy of the loosely academic sort, written by a man influenced by Ralph Waldo Emerson and by the 19th century English sociologist Herbert Spencer. “El Hombre Mediocre” begins this way: “When your visionary bow aims toward the stars and you unfold your wings toward such unreachable loftiness, eager to perfection and rebellious to mediocrity, you carry deep within, the mysterious spring of an Ideal.”
That was Daniel, it seemed to me — despite the mangled translation — in the throes of his life. [...]
H-2A is controversial. Some critics, advocating for an end to the program, believe there’s no point in tinkering with, modifying or reforming something that, in their opinion, is intractably unjust and inhumane at its core. Farmworker unions and labor advocates have compared it to slavery.
One problem is extortion. Odilon told me he sends home $15,000 during his stints of work in the U.S.; Daniel said he sends home $8,000. Given that there are hundreds of thousands of Mexican men in the United States each year on H-2A visas, it’s safe to assume that billions of dollars leave the American economy annually and end up in Mexico as personal earnings. Bad things can happen when money moves like that. For instance, labor recruiters who arrange for H-2A visas — and are paid for their services by American employers — might demand fees from people like Odilon and Daniel, whose accumulated dollars make them vulnerable generally. (That’s why they’re referred to in this story only by their first names.) [...]
Daniel said he’d agreed to be interviewed because he wanted to meet someone who had an important job. He wanted to talk about important things, he explained, with important people. What did he mean, I asked, by important people with important jobs? Lawyers, politicians, professors, journalists, people like that, he answered. He wanted to listen to how they spoke and ponder how their minds worked.
Daniel told me I should read his favorite book, “El Hombre Mediocre,” or “The Mediocre Man,” written in 1913 by an Argentine named José Ingenieros. It explained everything, he said ambiguously. Later I found a bad English translation. It’s philosophy of the loosely academic sort, written by a man influenced by Ralph Waldo Emerson and by the 19th century English sociologist Herbert Spencer. “El Hombre Mediocre” begins this way: “When your visionary bow aims toward the stars and you unfold your wings toward such unreachable loftiness, eager to perfection and rebellious to mediocrity, you carry deep within, the mysterious spring of an Ideal.”
That was Daniel, it seemed to me — despite the mangled translation — in the throes of his life. [...]
H-2A is controversial. Some critics, advocating for an end to the program, believe there’s no point in tinkering with, modifying or reforming something that, in their opinion, is intractably unjust and inhumane at its core. Farmworker unions and labor advocates have compared it to slavery.
One problem is extortion. Odilon told me he sends home $15,000 during his stints of work in the U.S.; Daniel said he sends home $8,000. Given that there are hundreds of thousands of Mexican men in the United States each year on H-2A visas, it’s safe to assume that billions of dollars leave the American economy annually and end up in Mexico as personal earnings. Bad things can happen when money moves like that. For instance, labor recruiters who arrange for H-2A visas — and are paid for their services by American employers — might demand fees from people like Odilon and Daniel, whose accumulated dollars make them vulnerable generally. (That’s why they’re referred to in this story only by their first names.) [...]
Meanwhile, there’s broad support in government and agribusiness for the H-2 program. Its growth, supporters argue, would allow for the deportation of workers here illegally while insuring a reliable and stable workforce. This, though, is just their short-term solution. “Ultimately,” Secretary of Agriculture Brooke Rollins said last July, “the answer on this is automation.”
by David Gutterson, Seattle Times | Read more:
Images: Ivy Ceballo/The Seattle Times
The Dot and the Swarm
I generally think I have done a good job anticipating the direction and pace of AI over the few years I have been writing this Substack, but I think I recently got something fairly large wrong. In the last year I have been posting about how I suspected that humans would have to approach working with agents as a manager, deciding how to delegate work to agents and specifying how those agents should be organized. I thought that getting agents to work effectively as a group would take careful construction, akin to building a company, and that this would take time to figure out.
Nope.
I fell prey to The Bitter Lesson, the hard truth, learned over and over again, that things that we thought required elaborate human rules and thinking can be solved with the brute force of better machine learning systems and more AI. The Bitter Lesson is everywhere among AI startups and companies adopting AI. A huge amount of effort went into building elaborate computer systems to feed AIs the right information at the right time, but AI systems have learned to seek out information themselves. The same thing happened to prompting. People built elaborate templates and chains of prompts that walked the AI through a task one step at a time. Then newer models turned out to be better at planning the steps themselves, and, as our research shows, planning steps have much less value. The history of the Bitter Le—
— you know what? I don’t really need to explain the Bitter Lesson, I asked Claude to do it in a music video. With one prompt, Fable wrote the lyrics and submitted it to Suno; Opus 5.5 did everything else using code alone without any image generation (How did Opus 5.5 pull this off? The Bitter Lesson tells you!). I gave no feedback at all.
Nope.
I fell prey to The Bitter Lesson, the hard truth, learned over and over again, that things that we thought required elaborate human rules and thinking can be solved with the brute force of better machine learning systems and more AI. The Bitter Lesson is everywhere among AI startups and companies adopting AI. A huge amount of effort went into building elaborate computer systems to feed AIs the right information at the right time, but AI systems have learned to seek out information themselves. The same thing happened to prompting. People built elaborate templates and chains of prompts that walked the AI through a task one step at a time. Then newer models turned out to be better at planning the steps themselves, and, as our research shows, planning steps have much less value. The history of the Bitter Le—
— you know what? I don’t really need to explain the Bitter Lesson, I asked Claude to do it in a music video. With one prompt, Fable wrote the lyrics and submitted it to Suno; Opus 5.5 did everything else using code alone without any image generation (How did Opus 5.5 pull this off? The Bitter Lesson tells you!). I gave no feedback at all.
by Ethan Mollick, One Useful Thing | Read more:
Image: AI and the author
Friday, October 9, 2026
Why Everybody Loves Glen Campbell
[ed. Nice tribute.]
But there was a catch—they weren’t his records. Campbell was just a member of the band, and the public didn’t know his name. Even so, what amazing company this young man kept: He now played guitar for Frank Sinatra, Elvis Presley, Phil Spector, Nat King Cole, Dean Martin, Bobby Darin, Sammy Davis Jr., the Everly Brothers, Doris Day, and dozens of other superstars.
He deserved every bit of his new-found success. Rock star Leon Russell, who also worked in these sessions, called Campbell the “best guitar player I'd heard before or since.” Songwriter Jimmy Webb later declared that Campbell could match up with “any guitar player in the world, from George Benson to Eric Clapton.” Eddie Van Halen was so desperate to get a guitar lesson from Campbell, that he traveled all the way to Arizona just to make it happen.
Campbell was that good.
There was, of course, a tiny problem. Glen didn’t know how to read music. So he sometimes asked the other musicians to play a new song without him the first time, so he could hear what was happening. But for the second run-through, Campbell was ready and contributing.
He played for union scale in those early days. That meant $65 for a three-hour session. In the early 1960s that was solid money. Back in Arkansas, Glen had picked cotton for $1.25 for every hundred pounds—that was more than a whole day’s work. Music was a much better option. And Campbell was in the studio constantly now. In 1963 alone, he showed up on 586 recorded songs.
But this young man hoped for more. Glen Campbell wanted to be a star. And he tried everything to make it happen. But nothing worked. [...]
Was there anything left to try? Yes, there was.
Around this same time, Glen Campbell rebranded himself as a surf musician, and got a choice gig with the Beach Boys. That’s quite a feat for a kid from a landlocked state. For a time, he actually replaced Brian Wilson as lead singer with the band. You might think this would turn him into a superstar, but fans wanted Brian back, and had no interest in celebrating his successor.
He played for union scale in those early days. That meant $65 for a three-hour session. In the early 1960s that was solid money. Back in Arkansas, Glen had picked cotton for $1.25 for every hundred pounds—that was more than a whole day’s work. Music was a much better option. And Campbell was in the studio constantly now. In 1963 alone, he showed up on 586 recorded songs.
But this young man hoped for more. Glen Campbell wanted to be a star. And he tried everything to make it happen. But nothing worked. [...]
Was there anything left to try? Yes, there was.
Around this same time, Glen Campbell rebranded himself as a surf musician, and got a choice gig with the Beach Boys. That’s quite a feat for a kid from a landlocked state. For a time, he actually replaced Brian Wilson as lead singer with the band. You might think this would turn him into a superstar, but fans wanted Brian back, and had no interest in celebrating his successor.
At this point, I would have given up and gone back to Arkansas. Campbell had made enough cash from those Beach Boys gigs to lease a Gold Cadillac, and could have returned home in grand style. But he was determined to try his luck one more time, and Capitol Records—then flush with cash, due to all those Beatles hits—still had money to burn.
That’s when they finally struck gold, or a gold record, to be more precise. Gentle on My Mind, released in August 1967, not only got Campbell on the charts but earned him two Grammy awards. He was now on a rapid trajectory to stardom, and would never look back.
During the next 34 months, Campbell released twelve more albums. This staggering output not only happened while he was concertizing and touring, but Campbell was now also acting in Hollywood movies. Adding to the craziness, he launched his CBS television show The Glen Campbell Goodtime Hour in 1969. [...]
That’s when they finally struck gold, or a gold record, to be more precise. Gentle on My Mind, released in August 1967, not only got Campbell on the charts but earned him two Grammy awards. He was now on a rapid trajectory to stardom, and would never look back.
During the next 34 months, Campbell released twelve more albums. This staggering output not only happened while he was concertizing and touring, but Campbell was now also acting in Hollywood movies. Adding to the craziness, he launched his CBS television show The Glen Campbell Goodtime Hour in 1969. [...]
Glen Campbell now saw his name in lights—as they say in show business. That was quite a change for someone who had grown up without any lights, or even electricity, in the family home. That was back in Billstown, a city so small that it didn’t show up on road maps.
His father was a sharecropper, and the family only ate what they grew or hunted—and sometime there wasn’t enough for the large Campbell family (which produced twelve children). There were days when Glen stayed inside his classroom during school recess, because it was cold and he didn’t own a coat. “We didn’t just endure poverty,” Campbell recalled, “we wore it.”
In his autobiography, Campbell tells of the time his father gave up all hope, and decided to commit suicide. The story is half a tragedy, and half a miracle:
He enjoyed every bit of it. Campbell built the first million-dollar house in Laurel Canyon—a 16,500 square-foot home on seven acres. He later bragged that he had “stayed in smaller three-story hotels.” But he could afford it—he was now making $50,000 or more in a single night.
But the enjoyment soon turned sour. “There were times when I was too drunk to recall what I did or with whom I did it,” he later admitted. He sometimes got bad write-ups in the press for the times he faltered on-stage.
In his autobiography, Campbell admits his failings, but can’t resist pointing out that the same journalists who criticized him were also drinking and using drugs, although nobody called them out. But the reality was that Campbell was under the bright lights, front and center, night after night—undergoing a kind of scrutiny no music critic ever faces.
The pressure was intense. “In the late 1960s and early 1970s, I didn’t have a career,” he later mused. “My career had me.”
His father was a sharecropper, and the family only ate what they grew or hunted—and sometime there wasn’t enough for the large Campbell family (which produced twelve children). There were days when Glen stayed inside his classroom during school recess, because it was cold and he didn’t own a coat. “We didn’t just endure poverty,” Campbell recalled, “we wore it.”
In his autobiography, Campbell tells of the time his father gave up all hope, and decided to commit suicide. The story is half a tragedy, and half a miracle:
Dad couldn’t bear the pressure of not being able to provide the bare necessities for his large family any longer. So he took a shotgun and headed into the woods. He loaded the weapon and was about to put the barrel in his mouth when he was distracted by a squirrel running down a tree branch. He shot and killed it. Almost instantly another came down, and another. My dad was crying and shooting his gun, and the family had eight squirrels for supper….Dad lived an additional forty years.The father was a survivor, but his son was destined for greater things and, like Johnny B. Goode, that little country boy could play the guitar. He got his first instrument at age four—a three-quarter size Sears Roebuck model that cost five dollars—and was gigging and going on the road while still a child. He joined his uncle’s band and they made the move to Albuquerque when Campbell was seventeen. Six years later he came to Los Angeles. And six years after that, Glen Campbell was a superstar.
He enjoyed every bit of it. Campbell built the first million-dollar house in Laurel Canyon—a 16,500 square-foot home on seven acres. He later bragged that he had “stayed in smaller three-story hotels.” But he could afford it—he was now making $50,000 or more in a single night.
But the enjoyment soon turned sour. “There were times when I was too drunk to recall what I did or with whom I did it,” he later admitted. He sometimes got bad write-ups in the press for the times he faltered on-stage.
In his autobiography, Campbell admits his failings, but can’t resist pointing out that the same journalists who criticized him were also drinking and using drugs, although nobody called them out. But the reality was that Campbell was under the bright lights, front and center, night after night—undergoing a kind of scrutiny no music critic ever faces.
The pressure was intense. “In the late 1960s and early 1970s, I didn’t have a career,” he later mused. “My career had me.”
by Ted Gioia, Honest Broker | Read more:
Image: uncredited; publicity photo
Mathematics Reels After New OpenAI Release
"If a human did this, it would be an instant Fields Medal, no questions asked.”
That’s how Alex Kontorovich, chair of the department of mathematics at Rutgers, responded to an A.I.-generated proof that OpenAI, the artificial intelligence giant, released on Tuesday.
Not all of them were quite so remarkable. But Martin Bridson, a mathematician at Oxford and the president of the Clay Mathematics Institute, which established the coveted Millennium Prize Problems, called OpenAI’s release “breathtaking.” [...]
If there had been any doubt that the leaps achieved by artificial intelligence this year would transform the discipline of mathematics, the tens of thousands of pages let loose on Tuesday have put that doubt to rest. The puzzles apparently conquered by OpenAI include crucial open questions in nearly every subfield of math, including algebra, number theory, theoretical computer science, mathematical logic and topology.
None of the proofs fully resolve any of the five remaining Millennium Prize Problems, which were designated at the turn of the century as a way of celebrating mathematics and its vast frontier of mysteries. But the trove of findings seemed to include results that make tangential or related progress on all of them. Perhaps the most significant result — the one Dr. Kontorovich was responding to — was a proof of the “quasi-Riemann hypothesis,” a conjecture related to the best known of the Millennium problems, the Riemann hypothesis, and a possible steppingstone of sorts to its resolution. [...]
‘Exciting for Many, Frightening for Others, Devastating for Some’
Talking to mathematicians after the release, it is clear that it is a highly conflicting environment, motivating a wild range of reactions. [...]
For Bryna Kra, a mathematician at Northwestern University, problem No. 145 on the list was of great interest — “the Rokhlin problem on mixing implying higher order mixing,” which answers a question from the 1940s. “The writing, however, makes it impossible to understand,” said Dr. Kra, former president of the American Mathematical Society, who is among a group of mathematicians who are proactively working to chart a course for the field through this period of “uncertainty and seismic change,” as she described it.
Overall, the results are “amazing,” Dr. Kra said, but now tens of thousands of pages of mathematical results need to be carefully checked. Typically checking and verification are done by peer review. “Peer review is gone in this process,” she said. “So, the question is, what kind of impact will these A.I.-generated results have? This is going to take a long time for the math community to digest.”
“Perhaps the most relevant issue is that by releasing this trove of material in this manner, the companies are in the process of destroying the ecosystem that made it possible for this work,” she said. “While this batch of papers has improved in the citations to the past literature, there are numerous indications that they have benefited from access to many arguments, only some of which are published. The mathematician who wants to make use of these amazing tools to advance their research instead risks giving their ideas away.”
The French mathematician Jean-Pierre Serre, an emeritus professor at the Collège de France in Paris — who has won two of math’s top prizes, the Fields Medal and the Abel Prize, and celebrated his 100th birthday last month — was naturally attracted to finding No. 46 in OpenAI’s list, which solves positively a conjecture he made almost 70 years ago.
But Dr. Serre is conflicted about doing math with A.I. He uses the technology for things like references, and, with the help of friends, for chasing down hunches about errors. “Mathematicians take pleasure in doing maths in two different ways: learning and finding new things,” Dr. Serre said. “Hence a conflict: A.I. increases the first pleasure and lowers the second one. The problem is it may lower the pleasure too much; that is especially serious for young mathematicians.”
by Siobhan Roberts, NY Times | Read more:
Image: Ian C. Bates
Also, if you were wondering: apparently they tried the model on about 8,000 problems. So, right now it “merely” solves ~5% of the longstanding open mathematical problems that it’s asked about, the problems that whole communities have spent years on, after a single 3-hour attempt on them. [...]
Update: As several people have pointed out, cryptography is a subfield that’s extremely conspicuous by its absence from OpenAI’s list of 376 papers! But my sources tell me that the AI companies have now started, gingerly and discreetly, investigating whether their latest internal models can break important cryptographic protocols and primitives. If they can, then it would certainly be nice to get ahead of things before the rest of the world figures out the same.
That’s how Alex Kontorovich, chair of the department of mathematics at Rutgers, responded to an A.I.-generated proof that OpenAI, the artificial intelligence giant, released on Tuesday.
And that single proof — worthy, in Dr. Kontorovich’s assessment on social media, of arguably the highest honor in the field of mathematics — was just one among more than 350 findings that the firm released that day, all at once.
Not all of them were quite so remarkable. But Martin Bridson, a mathematician at Oxford and the president of the Clay Mathematics Institute, which established the coveted Millennium Prize Problems, called OpenAI’s release “breathtaking.” [...]
None of the proofs fully resolve any of the five remaining Millennium Prize Problems, which were designated at the turn of the century as a way of celebrating mathematics and its vast frontier of mysteries. But the trove of findings seemed to include results that make tangential or related progress on all of them. Perhaps the most significant result — the one Dr. Kontorovich was responding to — was a proof of the “quasi-Riemann hypothesis,” a conjecture related to the best known of the Millennium problems, the Riemann hypothesis, and a possible steppingstone of sorts to its resolution. [...]
(As of Thursday, OpenAI had made a number of updates to the public online repository; withdrawing three papers and making fixes to several others. A spokeswoman said: “Where errors are identified, we will work to correct them promptly and withdraw papers if no fixes can be found. As with other research manuscripts, this is an iterative process.” )
‘Exciting for Many, Frightening for Others, Devastating for Some’
Talking to mathematicians after the release, it is clear that it is a highly conflicting environment, motivating a wild range of reactions. [...]
For Bryna Kra, a mathematician at Northwestern University, problem No. 145 on the list was of great interest — “the Rokhlin problem on mixing implying higher order mixing,” which answers a question from the 1940s. “The writing, however, makes it impossible to understand,” said Dr. Kra, former president of the American Mathematical Society, who is among a group of mathematicians who are proactively working to chart a course for the field through this period of “uncertainty and seismic change,” as she described it.
Overall, the results are “amazing,” Dr. Kra said, but now tens of thousands of pages of mathematical results need to be carefully checked. Typically checking and verification are done by peer review. “Peer review is gone in this process,” she said. “So, the question is, what kind of impact will these A.I.-generated results have? This is going to take a long time for the math community to digest.”
“Perhaps the most relevant issue is that by releasing this trove of material in this manner, the companies are in the process of destroying the ecosystem that made it possible for this work,” she said. “While this batch of papers has improved in the citations to the past literature, there are numerous indications that they have benefited from access to many arguments, only some of which are published. The mathematician who wants to make use of these amazing tools to advance their research instead risks giving their ideas away.”
The French mathematician Jean-Pierre Serre, an emeritus professor at the Collège de France in Paris — who has won two of math’s top prizes, the Fields Medal and the Abel Prize, and celebrated his 100th birthday last month — was naturally attracted to finding No. 46 in OpenAI’s list, which solves positively a conjecture he made almost 70 years ago.
But Dr. Serre is conflicted about doing math with A.I. He uses the technology for things like references, and, with the help of friends, for chasing down hunches about errors. “Mathematicians take pleasure in doing maths in two different ways: learning and finding new things,” Dr. Serre said. “Hence a conflict: A.I. increases the first pleasure and lowers the second one. The problem is it may lower the pleasure too much; that is especially serious for young mathematicians.”
by Siobhan Roberts, NY Times | Read more:
Image: Ian C. Bates
[ed. See also: New Math from OpenAI (DWAV); and, The Mathocalypse (Shtetl-Optimized):]
***
For those who are wondering: apparently, the AI model that produced all these wonders was not bespoke contraption of 10,000 agents burning millions of dollars worth of compute, as was used for example to construct a finite-time blowup for the Navier-Stokes equations. Instead, it was simply the latest internal OpenAI model—one that might be released to paying ChatGPT customers within the next couple of months, depending on the recommendations of OpenAI’s safety board! (My 9-year-old son: “Oh they definitely shouldn’t release that. If it could solve all those math problems, it can’t possibly be safe.”) Apparently they used about 3 hours of GPT-Pro level compute on average per problem solved.Also, if you were wondering: apparently they tried the model on about 8,000 problems. So, right now it “merely” solves ~5% of the longstanding open mathematical problems that it’s asked about, the problems that whole communities have spent years on, after a single 3-hour attempt on them. [...]
Update: As several people have pointed out, cryptography is a subfield that’s extremely conspicuous by its absence from OpenAI’s list of 376 papers! But my sources tell me that the AI companies have now started, gingerly and discreetly, investigating whether their latest internal models can break important cryptographic protocols and primitives. If they can, then it would certainly be nice to get ahead of things before the rest of the world figures out the same.
Thursday, October 8, 2026
Mamdani Is Showing the World How It’s Done
A few days before the leaders of the world streamed into the cavernous, marble-lined General Assembly Hall to deliver their speeches on the grave crises convulsing our combustible world, Mayor Zohran Mamdani offered his own, somewhat surprising solution to the problem of rising authoritarianism: filling potholes.
“If politics cannot deliver on the block where you live, why would you trust it to deliver anywhere?” Mamdani asked a gathering of mayors from big cities across the world, explaining his obsession with potholes, more than 200,000 of which have been filled across the city since he took office. “If you are making a New Yorker’s life a little smoother, so too you are making their journey to believe in government’s ability to deliver for them.”
It is commonplace to lament the ineffectual talking shop that the United Nations General Assembly has become, with its blustering speeches and endless unread communiqués. Wars of aggression are raging in Ukraine and Iran; Israel has carried out a genocide in Gaza; climate catastrophes impoverish and kill. In the face of these calamities, the organization appears powerless.
Today, the body’s soaring aims — to create a more peaceful, humane world that respects the sovereignty of each nation and the dignity of each human being — seem like a fantasy from another dimension. Citizens the world over are dissatisfied with their governments and in many places are souring on democracy itself. Can it be any surprise that this supranational body comes in for such contempt?
And yet Mamdani sensed an opportunity to reclaim some of its founding spirit. Racing across the city last week, he seemed to be offering an antidote to globe-spanning distemper. To restore faith in institutions, you must begin from the bottom up, starting with things as simple as potholes. Politics, as he put it, “has to translate to the pavement.”
New York City has long held a singular spot on the international stage, and its mayors have become global figures. But Mamdani, by dint of his youth, his improbable rise and his unusual background, brings a vivid symbolism to the role. African by birth, South Asian by heritage, American by choice, Mamdani in many ways embodies the cosmopolitan ideal at the heart of the United Nations’ founding.
World leaders lined up to spend time with him. He went to Queens with Ireland’s prime minister for a bingo game at an Irish American community center. He met with Brazil’s octogenarian leftist president and breakfasted with Spain’s dashing left-wing prime minister. Norway’s prime minister tagged along with him to visit a New York City classroom and talk about artificial intelligence in education.
Even President Trump couldn’t resist some time with the mayor. Last Monday he dropped by Gracie Mansion, spending about an hour behind closed doors with Mamdani — an eternity in presidential scheduling. Trump seemed in no rush to leave as the two took questions from reporters, beaming as he declared, “I want him to be a great mayor, and he’s certainly got potential, got great potential.”
Mamdani’s agenda, as befits a mayor, was mostly municipal. He visited a wastewater treatment facility with his counterparts from Helsinki, Rome and Barcelona; welcomed families moving into public housing with the mayors of Paris, Amsterdam and Greater Manchester; and handed out fliers to delivery workers explaining their rights with Mexico City’s mayor.
When he wasn’t hanging out with his fellow mayors, he celebrated the city’s victory in court against the delivery service DoorDash, wresting over $131 million from the company in a settlement for underpaying or not paying its workers. He also announced a new initiative to give out 70,000 free tickets to the Metropolitan Opera House, part of his push to make the city’s most exclusive cultural offerings accessible to everyday New Yorkers.
Mamdani’s many critics have tried to portray him as fixated on distant hobbyhorses, like arresting Israel’s prime minister, Benjamin Netanyahu, for war crimes in Gaza. But if anything, it was Netanyahu who seemed obsessed with the mayor. In his thundering speech at the United Nations last week, he called Mamdani an antisemite with “Hamas buddies.” He darkly name-checked the leftist streamer Hasan Piker and even castigated Mamdani’s wife for her likes on social media.
When I caught up with Mamdani by phone late last week, I asked him what he had hoped to get out of his first General Assembly. “We are a city that is home to hundreds of nationalities and even more languages,” Mamdani told me. “If any city is an illustration of what the United Nations principles hope to achieve, it is ours.”
But no matter how admirable those principles, the weakness of institutions like the United Nations in the face of the world’s spiraling chaos is an invitation to ask deeper questions, Mamdani said: “In a moment of a rising far right and the questioning of institutions that have otherwise been durable over many decades, it is also an opportunity to rethink what has been at the heart of those institutions, and the question of who they can serve and who they seek to serve.”
“If politics cannot deliver on the block where you live, why would you trust it to deliver anywhere?” Mamdani asked a gathering of mayors from big cities across the world, explaining his obsession with potholes, more than 200,000 of which have been filled across the city since he took office. “If you are making a New Yorker’s life a little smoother, so too you are making their journey to believe in government’s ability to deliver for them.”
Today, the body’s soaring aims — to create a more peaceful, humane world that respects the sovereignty of each nation and the dignity of each human being — seem like a fantasy from another dimension. Citizens the world over are dissatisfied with their governments and in many places are souring on democracy itself. Can it be any surprise that this supranational body comes in for such contempt?
And yet Mamdani sensed an opportunity to reclaim some of its founding spirit. Racing across the city last week, he seemed to be offering an antidote to globe-spanning distemper. To restore faith in institutions, you must begin from the bottom up, starting with things as simple as potholes. Politics, as he put it, “has to translate to the pavement.”
New York City has long held a singular spot on the international stage, and its mayors have become global figures. But Mamdani, by dint of his youth, his improbable rise and his unusual background, brings a vivid symbolism to the role. African by birth, South Asian by heritage, American by choice, Mamdani in many ways embodies the cosmopolitan ideal at the heart of the United Nations’ founding.
World leaders lined up to spend time with him. He went to Queens with Ireland’s prime minister for a bingo game at an Irish American community center. He met with Brazil’s octogenarian leftist president and breakfasted with Spain’s dashing left-wing prime minister. Norway’s prime minister tagged along with him to visit a New York City classroom and talk about artificial intelligence in education.
Even President Trump couldn’t resist some time with the mayor. Last Monday he dropped by Gracie Mansion, spending about an hour behind closed doors with Mamdani — an eternity in presidential scheduling. Trump seemed in no rush to leave as the two took questions from reporters, beaming as he declared, “I want him to be a great mayor, and he’s certainly got potential, got great potential.”
Mamdani’s agenda, as befits a mayor, was mostly municipal. He visited a wastewater treatment facility with his counterparts from Helsinki, Rome and Barcelona; welcomed families moving into public housing with the mayors of Paris, Amsterdam and Greater Manchester; and handed out fliers to delivery workers explaining their rights with Mexico City’s mayor.
When he wasn’t hanging out with his fellow mayors, he celebrated the city’s victory in court against the delivery service DoorDash, wresting over $131 million from the company in a settlement for underpaying or not paying its workers. He also announced a new initiative to give out 70,000 free tickets to the Metropolitan Opera House, part of his push to make the city’s most exclusive cultural offerings accessible to everyday New Yorkers.
Mamdani’s many critics have tried to portray him as fixated on distant hobbyhorses, like arresting Israel’s prime minister, Benjamin Netanyahu, for war crimes in Gaza. But if anything, it was Netanyahu who seemed obsessed with the mayor. In his thundering speech at the United Nations last week, he called Mamdani an antisemite with “Hamas buddies.” He darkly name-checked the leftist streamer Hasan Piker and even castigated Mamdani’s wife for her likes on social media.
When I caught up with Mamdani by phone late last week, I asked him what he had hoped to get out of his first General Assembly. “We are a city that is home to hundreds of nationalities and even more languages,” Mamdani told me. “If any city is an illustration of what the United Nations principles hope to achieve, it is ours.”
But no matter how admirable those principles, the weakness of institutions like the United Nations in the face of the world’s spiraling chaos is an invitation to ask deeper questions, Mamdani said: “In a moment of a rising far right and the questioning of institutions that have otherwise been durable over many decades, it is also an opportunity to rethink what has been at the heart of those institutions, and the question of who they can serve and who they seek to serve.”
by Lydia Polgreen, NY Times | Read more:
Image: David Dee Delgado/Reuters[ed. ... aka, Socialist Commie.]
The Secretive Bilderberg Group Is a Key to Understanding the Global Power Elite
Most countries in the western world operate under a practically unquestionable assumption that they are democratic, in the sense that there is a plurality of different interests in society at any given time and each of them has an equal opportunity of gaining significant political power.
The very concept of an elite class, with more influence than mostly everyone else, is deemed a relic of history. C. Wright Mills, author of The Power Elite, explained that these understandings have a certain romantic appeal, because a supposed balance of interests, without elite interference, makes “each of them really quite impotent, then no one set of higher circles and no manageable set of institutional arrangements can be held accountable for the events and decisions of our time.”
But in a world of constant conflict, economic strife, and political hypocrites, perhaps it is at least worth considering the other side of the argument for once?
In Mills’ conceptualisation of a power elite, he explained how a uniquely powerful grouping of economic, political, and military interests emerged after World War II that had the means to make decisions with significant consequences for the whole of American society.
The narrow and centralised group of individuals at the apex of economic, political, and military hierarchies constituted an elite with the technical and institutional means of power, which are beyond the reach of ordinary people. Far from conspiratorial, this was a group seeking to maintain its position of social power via the maintenance of its own coincidence of shared interests.
Indeed, Mills argued that “to accept either view – of all history as conspiracy or of all history as drift – is to relax the effort to understand the facts of power and the ways of the powerful.”
Writing in the 1950s, Mills’ conceptualisation of an elite was rooted in facts of power that existed, for the most part, within the nation state. One big exception to this was the Bilderberg Group that formed in 1954, which brought together elites from across the world into a transnational network. The group’s official website states that it was developed in response to opinions “expressed by leading citizens on both sides of the Atlantic.”
It fails to say how non-leading citizens, the general public, had a say in its organisation or operation. The World Economic Forum, Trilateral Commission, and Atlantic Council serve as similar transnational elite knowledge networks where individuals with shared, and in this case globalist, interests can work to further their own agenda. Since the era of globalisation moved political and economic power beyond any single country, so too have elites transcended the bounds of the nation. The problem, then, is that how do the masses, bound within borders, stand a chance of influencing the decisions that this transnational elite are making?
To think of institutions like the Bilderberg Group as inconsequential is to underestimate the role they play in constructing a global ideological consensus among elites that is then brought home and applied to the “nation.” Bilderberg allows elites to interact and acquire a shared education in how to think about the world in globalised terms. While it is an unofficial institution, in the sense that is it not legally constituted, the decisions it makes are subsequently endorsed by official agencies like the Organisation for Economic Co-operation and Development and national governments.
This is especially true of its neoliberal economic position. If there is a global ruling class, as critical political sociologists have identified, then to constrain political analysis to only the official institutions of the nation, and omit anything akin to Bilderberg, is a major misunderstanding of the ways of the powerful in modern times.
The Bilderberg Group’s powerful members come from across the world and represent the political, economic, military, technological, and ideological spheres of globalisation. There is absolutely no representation of alternative interests and ideological positions, such as small business, populists, socialists, or nationalists. [...]
Understanding the Bilderberg Group as the center of an elite knowledge network is far more accurate than any simplistic, and often naïve, claim to the contrary; especially from those that would malign it with the catch-all, ignorant label of “conspiracy theory.”
Knowing that the locus of political, economic, and military power now resides offshore helps to explain the pandemic of political disaffectedness sweeping across the planet. Some even see this as the emergence of a global civil war, which is surely the last thing the world needs at this very moment. Apparently, Bilderberg’s secretive nature is to ensure its participants can “speak freely in an environment of trust,” but precisely what would these people, who themselves claim a proclivity for transparency and democracy, want to speak about free from public discussion, criticism, and alternative points of view?
The public, rationally, can only surmise that it is nothing good. Unless politicians revert to assuming their proper role as national, and local, leaders, responsible for their own constituents’ needs as opposed to those of transnational corporations, the global crisis of democracy is surely set to continue.
by Stuart J. Hooper, The Wire | Read more:
Image: Reuters/File
[ed. The Press are typically excluded (see list of 2026 attendees here) and all discussions are conducted under the Chatham House rule (meaning you can use and report what you learned, but you cannot share who you got it from, nor their affiliations). See also: At Bilderberg’s bigwig bash two things are guaranteed: Kissinger and secrecy (Guardian).]
The very concept of an elite class, with more influence than mostly everyone else, is deemed a relic of history. C. Wright Mills, author of The Power Elite, explained that these understandings have a certain romantic appeal, because a supposed balance of interests, without elite interference, makes “each of them really quite impotent, then no one set of higher circles and no manageable set of institutional arrangements can be held accountable for the events and decisions of our time.”
In Mills’ conceptualisation of a power elite, he explained how a uniquely powerful grouping of economic, political, and military interests emerged after World War II that had the means to make decisions with significant consequences for the whole of American society.
The narrow and centralised group of individuals at the apex of economic, political, and military hierarchies constituted an elite with the technical and institutional means of power, which are beyond the reach of ordinary people. Far from conspiratorial, this was a group seeking to maintain its position of social power via the maintenance of its own coincidence of shared interests.
Indeed, Mills argued that “to accept either view – of all history as conspiracy or of all history as drift – is to relax the effort to understand the facts of power and the ways of the powerful.”
Writing in the 1950s, Mills’ conceptualisation of an elite was rooted in facts of power that existed, for the most part, within the nation state. One big exception to this was the Bilderberg Group that formed in 1954, which brought together elites from across the world into a transnational network. The group’s official website states that it was developed in response to opinions “expressed by leading citizens on both sides of the Atlantic.”
It fails to say how non-leading citizens, the general public, had a say in its organisation or operation. The World Economic Forum, Trilateral Commission, and Atlantic Council serve as similar transnational elite knowledge networks where individuals with shared, and in this case globalist, interests can work to further their own agenda. Since the era of globalisation moved political and economic power beyond any single country, so too have elites transcended the bounds of the nation. The problem, then, is that how do the masses, bound within borders, stand a chance of influencing the decisions that this transnational elite are making?
To think of institutions like the Bilderberg Group as inconsequential is to underestimate the role they play in constructing a global ideological consensus among elites that is then brought home and applied to the “nation.” Bilderberg allows elites to interact and acquire a shared education in how to think about the world in globalised terms. While it is an unofficial institution, in the sense that is it not legally constituted, the decisions it makes are subsequently endorsed by official agencies like the Organisation for Economic Co-operation and Development and national governments.
This is especially true of its neoliberal economic position. If there is a global ruling class, as critical political sociologists have identified, then to constrain political analysis to only the official institutions of the nation, and omit anything akin to Bilderberg, is a major misunderstanding of the ways of the powerful in modern times.
The Bilderberg Group’s powerful members come from across the world and represent the political, economic, military, technological, and ideological spheres of globalisation. There is absolutely no representation of alternative interests and ideological positions, such as small business, populists, socialists, or nationalists. [...]
Understanding the Bilderberg Group as the center of an elite knowledge network is far more accurate than any simplistic, and often naïve, claim to the contrary; especially from those that would malign it with the catch-all, ignorant label of “conspiracy theory.”
Knowing that the locus of political, economic, and military power now resides offshore helps to explain the pandemic of political disaffectedness sweeping across the planet. Some even see this as the emergence of a global civil war, which is surely the last thing the world needs at this very moment. Apparently, Bilderberg’s secretive nature is to ensure its participants can “speak freely in an environment of trust,” but precisely what would these people, who themselves claim a proclivity for transparency and democracy, want to speak about free from public discussion, criticism, and alternative points of view?
The public, rationally, can only surmise that it is nothing good. Unless politicians revert to assuming their proper role as national, and local, leaders, responsible for their own constituents’ needs as opposed to those of transnational corporations, the global crisis of democracy is surely set to continue.
by Stuart J. Hooper, The Wire | Read more:
Image: Reuters/File
[ed. The Press are typically excluded (see list of 2026 attendees here) and all discussions are conducted under the Chatham House rule (meaning you can use and report what you learned, but you cannot share who you got it from, nor their affiliations). See also: At Bilderberg’s bigwig bash two things are guaranteed: Kissinger and secrecy (Guardian).]
Wednesday, October 7, 2026
AI Will Be a Boon for Savers
Sometime in the next few years, an ordinary American will give a simple instruction to an AI assistant: “Make sure my money is always working.” From that moment on, his cash may never sit still again.
The agent will sweep idle savings into whatever insured account pays the most. It will find high-paying savings accounts and ladder certificates of deposit without the owner’s learning what laddering means. This is agentic banking—software that displays your finances and acts on them—and the banking industry isn’t prepared for how fast it’s coming.
A staggering share of bank profitability rests on money that doesn’t move. This is true across the board, not only for big banks.
Americans hold roughly $10.4 trillion in savings and other liquid deposits. The Federal Deposit Insurance Corp.’s national average savings rate—the deposit-weighted average across every insured institution—is a mere 0.38%. At that rate, the system pays savers about $40 billion a year. The best nationally available savings accounts tracked by Bankrate pay around 4%. The same balances at that rate would earn more than $400 billion.
Even extremely conservative assumptions—half those balances stay put for good reasons, and top rates compress as money moves—leave Americans forgoing well over $150 billion a year. Call it the laziness tax: a quiet, regressive wealth transfer from inattentive consumers to the banking system.
AI agents will largely abolish it. The inertia that keeps a saver from spending 20 minutes to earn $1,000 disappears when the 20 minutes do. Every account becomes a shopped account. For consumers, this is unambiguously good: the largest, fastest transfer of value from financial institutions to households in generations, accomplished by software rather than regulation.
An agent working for a consumer can also flag junk fees, catch unauthorized charges, decode fine print and steer its owner away from predatory products. It can help accomplish, one household at a time, much of what the Consumer Financial Protection Bureau was created to do through enforcement. The regulatory challenge will be ensuring that agents’ actions stay aligned with consumers’ best interests.
But AI agents will strain the system that makes lending possible. Banks will be funding long-duration loans with money that can leave in 30 seconds. The Silicon Valley Bank crisis of March 2023 offered a preview of what can happen when technology accelerates money: Tens of billions of dollars left one bank in a day. Agents automate and accelerate that speed. Liquidity models built on human sluggishness will become a distant memory.
To survive when AI moves money automatically, banks will need to change how they do business.
by John Delaney and Bill Foster, Wall Street Journal | Read more:
Image: Getty Images
The agent will sweep idle savings into whatever insured account pays the most. It will find high-paying savings accounts and ladder certificates of deposit without the owner’s learning what laddering means. This is agentic banking—software that displays your finances and acts on them—and the banking industry isn’t prepared for how fast it’s coming.
A staggering share of bank profitability rests on money that doesn’t move. This is true across the board, not only for big banks.
Even extremely conservative assumptions—half those balances stay put for good reasons, and top rates compress as money moves—leave Americans forgoing well over $150 billion a year. Call it the laziness tax: a quiet, regressive wealth transfer from inattentive consumers to the banking system.
AI agents will largely abolish it. The inertia that keeps a saver from spending 20 minutes to earn $1,000 disappears when the 20 minutes do. Every account becomes a shopped account. For consumers, this is unambiguously good: the largest, fastest transfer of value from financial institutions to households in generations, accomplished by software rather than regulation.
An agent working for a consumer can also flag junk fees, catch unauthorized charges, decode fine print and steer its owner away from predatory products. It can help accomplish, one household at a time, much of what the Consumer Financial Protection Bureau was created to do through enforcement. The regulatory challenge will be ensuring that agents’ actions stay aligned with consumers’ best interests.
But AI agents will strain the system that makes lending possible. Banks will be funding long-duration loans with money that can leave in 30 seconds. The Silicon Valley Bank crisis of March 2023 offered a preview of what can happen when technology accelerates money: Tens of billions of dollars left one bank in a day. Agents automate and accelerate that speed. Liquidity models built on human sluggishness will become a distant memory.
To survive when AI moves money automatically, banks will need to change how they do business.
by John Delaney and Bill Foster, Wall Street Journal | Read more:
Image: Getty Images
[ed. Possible. But it'll take a heavy lift to get people to hand over their financial accounts (and security protections) to AI. What about criminal AIs that exploit that?]
Tuesday, October 6, 2026
The Canton Spirituals
[ed. Yeah... come back in a few years (like maybe three or four for every year you fucked up). Then maybe you'd have enough perspective to preach how you'd screwed everyone over while "making a lot of mistakes". Contrition stories will bloom like weeds after the mid-terms. Guaranteed.]
[See also: The Trump Administration is a Cesspit of Corruption (Krugman). But we already knew that.]
Subscribe to:
Posts (Atom)