His predictions (while often wrong) are at least somewhat justified and backed up by genuine scoops and original research. Altman goes on podcasts and talks about building Dyson spheres for energy.

For agentic work, but especially for web search, 3.6 Flash has an important leg up, its fast speed, that no other model comes close to matching. I guess nobody pays attention to it because it's not the one big flashy number that you compare to other models.

New 3.5 Flash model is clearly quite optimized for speed. People on this site just look at benchmarks, and underestimate how vital speed is to user experience. No other free model comes close in terms of speed and intelligence, so it seems like the correct strategy to slowly win the mass market, as almost everyone will form their experience of AI through its free version.

The answer to your question is “because the market isn’t big enough”, not because it doesn’t work. Why would knowing about 2019 internet memes help you in any way at coding?

One need only look at the resume filtering process, a once manual bias that has now been codified into algorithmic bias with AI. A degree from a good school boosts your chances immensely, and other facets such as coursework don't matter much.

If you have ever seen someone filter applicant resumes, you will understand instantly. There are too many, you have to filter them somehow and the allure is irresistible.

Meanwhile Greg Brockman is worth all the Anthropic founders PUT TOGETHER, he and his wife are the single largest donors to Trump, and he and Altman have formed a board full of sycophants and stolen a non-profit. When Altman was fired, they manipulated their morally bereft, money-hungry employees to get their own way. They have reneged on every single promise they've made as soon as it's inconvenient.

Why do I care about the models again?

It is not clear that running one's own datacenter is a competitive advantage. Why do you think OpenAI can handle that?

His biggest competitor asks people in the interview process if they’d be willing to give up their Anthropic stock for the good of society.

Surely you cannot just close your eyes and say they’re the same. Don’t allow evil to roam free under the guise of merely “imperfection”.

The reason media training exists is to win over people like you.

I know people high up at OpenAI, I'm quite sure Altman etc. don't care either; the only difference is, they've been trained well to hide it.

I don't understand the hate Gemini gets on this site. It is quite well-rounded, it is often better at tasks without benchmarks, like translation, but nobody gives credit where credit's due.

I can't ever forget that blog post where OpenAI advertised ChatGPT's improved image understanding, showing a labelled motherboard with several errors -- then someone posted Gemini's work on the exact same picture, with every component labelled perfectly. That doesn't show up on benchmarks but it matters to real people.