I remember buying into OpenAI's claim that I had Aristoteles in my pocket, and at first I was amazed. Then one day I realised it wasn't Aristoteles at all, it was my little brother Timmy. From that day on, whenever Timmy says something, I reply with: Nope, No, Again, Wrong Timmy, It doesn't work, and sometimes, Do better Timmy!
Timmy relies on patterns learned from huge amounts of text, so his answers are always relative and unpredictable.
Now AI companies are shifting how they talk about intelligence. They admit LLMs can't truly reason with the current architecture, so intelligence is being sold as the ability to solve problems using learned patterns, not actual reasoning. That's a big downgrade from the original dream of human-level thinking and AGI.
Still, I'm impressed with little Timmy and I know he will keep growing and getting smarter. But I don't see him solving climate change just yet.
Newton was obsessed with the idea that matter could be transformed, which is basically what alchemy is about. He thought metals like lead could be changed into gold if you uncovered the hidden principles of nature. In a way, he was chasing the same dream that much later became real with nuclear reactions.
I'd argue that shamelessness and countersignalling are different things but share the same foundation: confidence.
Shamelessness is acting without embarrassment and countersignalling is deliberately downplaying because you're so confident you don't need to prove yourself.
Using the example from the article, another person who comes to mind besides Paris Hilton is Trump. He uses countersignalling as a strategic tool, and sloppiness as a Swiss knife. The followers of both Paris and Trump interpret that sloppiness as confidence and authenticity, which is why it's so effective. And to pull off being deliberately sloppy, you need to be shameless.
I don't think it is that deep for Trump. His sloppiness is great for the media because his choices lead to endless content. He is great for the unofficial media because everything he does is meme worthy. It is no wonder he and Musk teamed up, both their successes come from following the same strategy.
This is actually a great time to launch a product for 13-18 year olds who'll be looking for something else to do on their phones during the 6 hours they used to spend watching porn.
I'm thinking of an AI tool that automatically dresses people in porn videos, turning them into 12+ content. The Pro+ plan could even replace the moaning with actual dialogue. That way they can keep watching their favourite movies without needing to verify their age or use a VPN.
The only thing I need for VCs to invest is to put the term "10x" on my website.
In other words, if they can't watch porn, they'll probably start practicing it. This is good news! Population collapse will stop, economy will grow, housing prices will soar, pension funds will have new contributors in 20 years, etc. etc.
> the report does not mention the leading causes of death for American children, which are firearms and motor vehicle accidents.
I'm honestly tired of seeing all these AI wannabes bragging on Twitter and LinkedIn about how they've built a tool that does "10x better" than something else, and how they only hire people willing to work 80 hours a week, with no life and no need for money. They're trying to normalise tech sweatshops because their VCs told them that being a psychopath is the fastest way to get a return on investment. These founders are so driven and so dumb that I have more respect for flat-earth conspiracy theorists than for this new wave of brainless, heartless founders.
And I don't buy into their Mother Theresa's story either, that they'll exploit people for decades, cash out billions, and then magically turn into philanthropists to fix all the harm they caused.
This is the time to use technology to solve real problems. That's the mission, that's why we're here. The leading causes of death for American children are firearms and motor vehicle accidents. The Volvo engineer who invented the seatbelt is a hero because he saved millions of lives.
We need to stop these shallow founders from destroying the reputation of our industry.
I used GPT-5 as my personal lawyer the other day. I uploaded contracts and agreements, and the number of things it got wrong was mind blowing. It made me look like a complete amateur with the emails it drafted. GPT-5 is extremely dumb and incompetent when it comes to understanding real-world problems and offering solutions. It wasn't even at the level of a junior lawyer, it was ten times worse. I was honestly shocked by the results.
When it comes to programming, I have to keep replaying with "Nope", "No", "Again", "Wrong", "It doesn't work" and a couple of times "Do better" before it finally produces something complex that actually works.
With coding at least I know when it's wrong. The real problem is when I don't.
But if you let it run the tests, it’ll just do that on its own (or delete the tests, but usually it won’t), which is very useful and seems to match the article’s sentiment.
Since LLMs rely on patterns learned from large amounts of text, the results are relative. If the training data contained more Rust repos, that could explain why it feels stronger in Rust.
The way AI companies talk about "intelligence" now is shifting. They admit LLMs can't truly reason with the current architecture, so intelligence is being framed as the ability to solve problems using patterns learned from text, not reasoning on their own. That's a big downgrade from the original idea of AI reaching human-level thinking and developing AGI.
Also, my understanding is that since Microsoft invests in Copilot, it doesn't want ChatGPT to get better at coding. Instead, it wants it to get better at being a lawyer.
Whenever people ask Google, Meta, Microsoft, or Twitter why their tools are polarising friendships, doing the Nazi salute, or stealing copyrighted material, they say: "Hey, we didn't know our AI was doing that. We didn't even know it could do that."
Well, next time they find Chinese malware on the phone of the NSA director, China should just say: "We didn't know Bruce AI was hacking phones! We only trained it to recommend products on Alibaba."
I guess what I'm trying to say is, with AI becoming so popular and quantum computers around the corner, either we all get serious about the truth or no one will. Because big tech companies in the US are lying through their teeth, about the AI they're building for military use, about the satellites they're launching, about everything. Then they tell the world, "they're bad, we're good." That doesn't work anymore. More and more developing countries are partnering with China, and perceptions are shifting. While the US focuses only on the cyber war, China is winning the tech, cyber, and diplomatic wars.
For me, the turning point was Covid-19. While the US was selling vaccines to wealthy countries, China was donating them to developing ones and saving lives. It shipped over a billion doses across Asia, Africa, and Latin America. In my country, China is seen as the good guy now.
