I'm really interested to know how you determine what is or isn't a bot comment, because I have showdead enabled, and I can guarantee you there is absolutely no difference in writing style or ideas between the dead comments and the other ones.
No, it's not. Google hasn't been competitive at all. Allegedly their latest model closed the gap somewhat, but what I heard is that it doesn't hold up too well outside of benchmarks.
For OpenAI, they have not been in a good place for the last two months. I switched from Astra to Opus 5.5 and it's been a clear upgrade, nicer to use, and much more generous with the usage limits on top.
This has nothing to do with rationalization. Opus 5.5 is just the better model, and that's a wide consensus. In the past I used the $200 subscription for both Anthropic and OpenAI, nowadays I downgraded Codex to $20.
From the comments on this release, it appears OpenAI rushed something barely usable out of the door in an attempt to stop bleeding customers to Anthropic. This would check out with the reports back then that the 6.1 Astra release was cancelled because of poor alignment and lack of usability as a coding agent.
I would not be surprised if the issue is still present to some degree in this 6.2 release.
It's possible they're inaccurate, but let's think about what that implies:
The New York Times, arguably the most prestigious newspaper of record in the United States, published an article more-or-less alleging that a man worth hundreds of billions of dollars, was having an affair with a person that the US government had at least some reason to believe was an intelligence agent working on behalf of the People's Republic of China, all without any real ability to back that story's publication up.
As a result, this man was forced to resign from the company he founded, had embarrassing details about his personal life brought to light, and the company's market capitalization went down by over a trillion dollars, causing losses to shareholders. There has been an unprecedented loss of reputation and money as a direct result of the Times' coverage of the story.
If the Times has no real backing for what they published, they have opened themselves up to the single largest libel suit of all time. It'd likely be the end of them as a publication. Most of the journalists involved in the story would never find work again.
The alternative to the above is that the reporting was true so far as the NYT could verify and that there are real questions that need to be asked about what exactly Alice Huang was doing at MegaSpeed.
I think it's rather a problem with the training pipeline incentivizing correctness and endlessly defensive code over simplicity and maintainability of the implementation.
The problem with context I see primarily in mixing up user intent and hard requirements with random half-correct notes some other agent left in the comments.
So yes, obviously he would be instructing the model to improve that port autonomously.
reply