DanielHall

Just to achieve a great ideal: MEGA Make Europe Great Again.

What a clickbait title. I thought Fable was no longer included in the Max plan.

Making a fuss over nothing.

Thank that 'Super'AILab supervisor from ScaleAI, Alexander Wang; this guy is really hilarious. He directly turned Meta into a Chinese company (just like how ScaleAI exploits its employees), and so far, I haven't seen him deliver anything that matches his annual salary. Considering that what he does is AI infrastructure, even cheap-to-the-point-of-ridiculous cheap labor for training data annotation. I don't think he's suitable for this kind of big-picture AI research.

These small models, having been fine-tuned for the test, achieve frighteningly high scores, yet perform abysmally in real-world scenarios.

the specific complaints are fair (gatekeeper friction, tahoe/liquid glass, the UK age check thing) but I always wonder who these "X just lost me" posts are for. Apple doesn't care. They care about churn numbers, not individual blog posts. also like... credit card as age verification isn't an Apple invention, Google and Meta does the same thing. And "your account region should match where you live" has been the rule forever. If you were running a cross-region account that whole time you were basically on borrowed time. Sucks when it finally breaks but it was never really supported. idk, as a "here's the stuff that finally pushed me over" post it's fine. the "$BIGCO just lost me" framing is what makes it eye-roll-y. they didn't lose you, you left. there's a difference and it matters because one implies they should care.

OpenClaw is underwhelming, and its founder is basically a hype machine.

Anthropic is set to become the second-strongest legal department in North America

Comparing open-source models like Qwen against Anthropic’s models is absolutely foolish. First of all, Anthropic has never disclosed the actual parameter count or architecture of their models. Second, it’s well known that these open-source models more or less distill from other models and use MoE, which allows them to run at much lower computational costs. Using Qwen as a comparison point only proves the blog post author is foolish. The article devoted such a large portion to discussing Qwen on OpenRouter, I find it hard to believe.

Another rather peculiar point: why did the Department of War reject Anthropic and label it a supply chain risk entity, despite its "prohibited content" policy being almost identical to OpenAI's stated policy in its announcement, yet award the contract to OpenAI?

A bit surprised, the first one released wasn't Sonnet 5 after all, since the Google Cloud API had leaked Sonnet 5's model snapshot codename before.

I encountered the same situation too; Claude has 'become lazy'.