Viewing profile — heaney-555

I hope this is at least done with GPT Live or equivalent, not some cheap shitty model. (but in general, it sounds like a bad idea, it's too early for this!)

The free tier of ChatGPT is powered by GPT-5.5 Instant, a non-reasoning version of GPT 5.5.

This is the problem with the genAI image hysteria. As the models improve, it becomes a fake outrage, since the ability to distinguish human creations from AI becomes impossible.

Technically yes, but the free usage equated to single-digit messages per day before the auto-downgrade. With GPT-5 it was just 1 message per day. Perhaps I should have said "proper…

Giving free ChatGPT users access to reasoning (the 'Think' toggle) will have a broader impact on the world than every new paid model and coding agent combined.

Sure, but the question is why either of those things happened, given Google's immense resources and talent. The fact that OpenAI, Anthropic, SpaceXAI and 3 different Chinese compan…

>In broad strokes, the Netherlands is a small-scale, low-tech society. Small-scale in that the city hall here has just one window open to queue at, whereas Shanghai has hundreds of…

The Gemini 3.5 Pro delay has been catastrophic for Google. I'm really curious what has gone on behind the scenes there. But missing out right as AI coding agents become genuinely d…

Waymo does 4 million rides per month. Uber does 1 billion, for reference. There's still a long, long way to go. Waymo has proven it can operate a high-priced service in a handful o…

So the goalposts have moved all the way to "LLMs can't do what Albert Einstein did"?

https://openai.com/index/introducing-gpt-live/ GPT-Live, which is coming to the API soon, responds instantly while reasoning in the background. So it can say "Hold on, I'll look th…

It's far far too early to say whether it turned out to be wrong. Tesla also has a robotaxi service in Dallas, FWIW, in testing.

https://medium.com/the-generator/a-waymo-vehicle-hit-a-kid-b... "As the Waymo approached the double-parked SUV, a child ran out from behind the SUV and into the roadway, directly i…

>The class of small models, with limited to no reasoning What? GPT-4.1 was not a small model! And why wouldn't you use reasoning? You're of course going to see poor results when yo…

>This seems to be the end of the road for LLM's This is an amazingly ignorant thing to say given the current pace of progress.

>We study a frontier LLM in its purest inference regime - a single generation pass over a prompt containing the full training and test data, with no tools, no agentic scaffolding, …

Have you seen Midjourney 8.2? Pretty soon, you will not be able to tell. What you really mean is "obvious AI-generated images".

Being late is different from being wrong.

The best tip is to provide a concrete example, or ideally 3+ examples, of the kind of writing/explanations you personally think is great. It's extremely good at matching when provi…

>Reading AI output is extra effort. It's verbose, frequently contains all too plausible nonsense, and is increasingly jargon dense. You can set up custom instructions to be less ve…

>Is there not value in trying to understand what the strengths and weaknesses of the technology are? There is, but having that discussion without a discussion and understanding of …

Do you seriously not see the difference, or are you being intentionally obtuse? Plus, no, multiple of the problems Astra solved do have real-world applications in computer science,…

You couldn't. PhDs have been working on these problems for decades. It wasn't for lack of trying that none of them could figure these solutions out!

These is mostly mathematics, not science, and they link the paper in the post: https://cdn.openai.com/pdf/74c24085-19b0-4534-9c90-465b8e29a... Mathematicians will tear it to pieces…

[dead]