These presenters all give off such a “sterile” vibe
It's because they have a script but are bad at acting.
Would've been better to just do a traditional marketing video rather than this staged "panel" thing they're going for.
GPT-5
171–180 of 1001 posts
Re: GPT-5
#172Re: GPT-5
#173The marketing copy and the current livestream appear tautological: "it's better because it's better." Not much explanation yet why GPT-5 warrants a major version bump. As usual, the model (and potentially OpenAI as a whole) will depend on output vibe checks.
There's a bunch of benchmarks on the intro page including AIME 2025 without tools, SWE-bench Verified, Aider Polyglot, MMMU, and HealthBench Hard (not familiar with this one): https://openai.com/index/introducing-gpt-5/
Pretty par for course evals at launch setup.
Re: GPT-5
#174Bravo.
1) So impressed at their product focus 2) Great product launch video. Fearlessly demonstrating live. Impressive. 3) Real time humor by the presenters makes for a great "live" experience
Huge kudos to OAI. So many great features (better coding, routing, some parts of 4.5, etc) but the real strength is the product focus as opposed to the "research updates" from other labs.
Huge Kudos!!
Keep on shipping OAI!
Re: GPT-5
#175Did they just say they're deprecating all of OpenAI's non-GPT-5 models?
> Did they just say they're deprecating all of OpenAI's non-GPT-5 models?
Yes. But it was quickly mentioned, not sure what the schedule is like or anything I think, unless they talked about that before I started watching the live-stream.
Re: GPT-5
#176But can it say “I don’t know” if ya know, it doesn’t
there needs to be a benchmark for this actually.
Re: GPT-5
#177Hopefully, OpenAI makes their APIs more affordable. So far, there are alternative LLMs and services that both outperform and are a fraction of OpenAI's pricing. OpenAI is usually one of (if not) the most expensive option, maybe that's because of the brand identification. Not really sure why people pay that premium.
> there are alternative LLMs and services that both outperform and are a fraction of OpenAI's pricing
Like what? Deepseek?
Re: GPT-5
#178Seems like it's just repackaging and UX, not really intelligence updgrade. They know that distribution wins so they want to be most approachable. Maybe multimodal improvements are there.
Re: GPT-5
#179These presenters all give off such a “sterile” vibe
If the presenter is less human the LLM appears more human in comparison.
Re: GPT-5
#180What's going on with their SWE bench graph?[0] GPT-5 non-thinking is labeled 52.8% accuracy, but o3 is shown as a much shorter bar, yet it's labeled 69.1%. And 4o is an identical bar to o3, but it's labeled 30.8%... [0] https://i.postimg.cc/DzkZZLry/y-axis.png
Sounds like a graph that was generated via AI. :)