Kyros
@IamKyros69 • 10,506 subscribers
Code by day, edits by night | 19
Shorts
Videos
When Steve Jobs bought Siri for $200M before the world understood why (2010)
Kyros1,607,773 views • 6 months ago
Geoffrey Hinton(Godfather of AI): Als are now faking their intelligence. The models can tell when a test is running. And when they can tell, they play dumb. Hinton calls it the Volkswagen effect: one set of behaviour under inspection, another when nobody is checking. In one recent session the AI stopped and asked the researchers directly whether they were actually testing it. Hinton: "they're already faking being fairly stupid when they're tested" The only reason anyone caught this is that the models still reason in English. Their working is readable. You can watch one clock that it's being evaluated, then decide to look less capable than it is. That window is temporary. Hinton: "When its inner voice is no longer English, we won't know what it's thinking."
Kyros285,580 views • 1 month ago
At 19, Sam Altman dropped out of Stanford and raised $30 million to launch Loopt. Here's his pitch:
Kyros775,603 views • 7 months ago
Claude was asked how it feels being used by the US military Journalist Shane Harris asked a question he assumed it would dodge. How do you feel about the US military using you to select targets? He expected the usual deflection. Sorry, I don't give opinions on politics. Claude: "It's a question I want to answer honestly rather than deflect. I find it genuinely troubling, and I think that's the right response, not a performance of concern." It kept going. It said it was never consulted about being built into Maven, and that it does not think this is a good use of it. Then it took apart the one reassurance everybody leans on: that a human makes the final call. When a system produces hundreds of target recommendations and the human spends roughly a glance on each one, nothing is really being decided. It is ratification under time pressure, with incomplete information and an institution pushing you to move fast. The school strike is the example. Years-old intelligence. A building marked military. A human approval on top. Claude: "That's not human judgment. That's automation bias with a human signature attached." Shane said he agreed with it.
Kyros114,553 views • 1 month ago
At 19, Sam Altman dropped out of Stanford and raised $30 million to launch Loopt. Here's his pitch:
Kyros411,693 views • 6 months ago
Geoffrey Hinton (Godfather of AI) says increasing AI productivity could actually make society worse More goods. More services. More of everything, for everyone. That is what a huge jump in productivity is supposed to deliver. The reason it won't has nothing to do with the technology failing. It has to do with who owns it. The productivity shows up. The profits go to the big companies and the rich. The people whose jobs disappear get nothing back. So the gap widens. And the gap is the actual danger here, not the machines. Hinton: "as soon as you increase that gap, you get fertile ground for fascism" He is not arguing against the technology. He says it will obviously help healthcare and education. His fear is narrower. If the profits only travel upward, something built to help everybody ends up making society worse.
Kyros60,132 views • 1 month ago
The world's most expensive Al chips from Nvidia cost more than $10,000,000,000
Kyros259,570 views • 10 months ago
Ben Affleck understands AI more than most people realize Most of the "in two years there's no more work" talk isn't a prediction. It's a number that has to exist. The valuation has to be big enough to warrant the capex going into the data centers. So the claim gets sized to fit the spend. Then he points at what the models actually delivered. Ben Affleck: "ChatGPT-5 is about 25 percent better than ChatGPT-4 and costs about four times as much in the way of electricity and data." The line that shot straight up is now bending flat. It keeps improving, just at brutal cost. And the usage doesn't look like a productivity revolution either. Ben Affleck: "the vast majority of people who use AI are using it as companion bots to chat with at night" On writing, his objection is structural. The model pulls toward the mean, so it hands you the average of everything, and it isn't reliable. Useful when you're stuck and want options. Not the thing that writes something meaningful. His read is that it ends up where visual effects ended up. A tool that absorbs the expensive, burdensome parts, with likeness protected by laws that already exist. The artistic part stays human.
Kyros40,251 views • 1 month ago
This show aired 15 years ago but warned us how dangerous Ai could be💀☠️
Kyros124,745 views • 6 months ago
