simonw

21 ms·

451.
by simonw 2mo ago
They say that they are yet to see: > "Our AI grew sentient, recognised the climate crisis, and shut our operations down." I guess they missed the whole https://www.snitchbench.com/ thing where it turns out many
452.
by simonw 2mo ago
You mean in the comment where I suggested that I didn't think it was a very good product? I posted this comment because I didn't know what the author meant by ONA and had to look it up myself.
453.
by simonw 2mo ago
What the heck? What product was I advertising here?
454.
by simonw 2mo ago
When this article talks about ONA it means https://ona.com/ - a cloud agent service that was acquired by OpenAI a couple of months ago. (I wouldn't suggest basing my evaluation of the entire field of coding agents arou
455.
by simonw 2mo ago
There's a big gap between being able to ask questions about something and understanding something. The more things you understand, the higher the chance you'll spot a situation to use them in the future. I think the best innovat
456.
by simonw 2mo ago
What are your favorite examples of him misleading his readers?
457.
by simonw 2mo ago
https://github.com/nsrht/time-travel-sqlite-debugger/blob/ma... works by checking the mtime on the database file (and -wal file) once per second and creating a backups/unixtime_database.sqlite file if an
458.
by simonw 2mo ago
Yeah, any time I see someone accuse Gruber of being universally positive about Apple I think they can't have been reading his work very often. Something Is Rotten in the State of Cupertino: https://daringfireball.net/20
459.
by simonw 2mo ago
The bigger issue here is that he mislead (I think lied to, personally) John Gruber to the point that Gruber had to issue his first retraction in 25 years: https://daringfireball.net/2026/08/retraction_app_store_rej
460.
by simonw 2mo ago
Yeah, the lobbying is gross.
461.
by simonw 2mo ago
I think part of the problem is that they had been running that Artifactory configuration previously without any problems, and it gave them a false sense of security. Similar thing happened with the UK AISI - they got caught out because the
462.
by simonw 2mo ago
> I learned about this technique in 2007, then tried writing it up in 2015. I realized that I didn’t understand it enough to be able to explain it. I studied it off and on in 2016, 2018, 2019, 2022, 2024, and 2026. I abandoned and restar
463.
by simonw 2mo ago
They key lesson I've picked up from the past ~4 months is that models are now good enough that, if there's a security hole, they'll brute force their way into finding it. The only solution that makes sense to me is for defend
464.
by simonw 2mo ago
What are my goals here?
465.
by simonw 2mo ago
Because nobody has figured out how to make that work 100% reliably yet. The current approach is to use delimiters that are special tokens that can't be represented in regular text: https://github.com/openai/harmony
466.
by simonw 2mo ago
I'm fed up with compaction. I want my agent to get compacted but also retain full access to the prior conversation via search and tool calls - I want it to know "the requirements for X were discussed in detail previously in conver
467.
by simonw 2mo ago
> But if the solution hadn’t been in the knowledge it would not have been able todo it. Part of the solution involved discovering two separate zero-day vulnerabilities in Artifactory, so saying the solution must have "been in the kn
468.
by simonw 2mo ago
Because it's a crime. Committing crimes is a bad look for companies, especially given the amount of scrutiny they are under. Would you deliberately commit computer crimes when the Trump admin yoinked Fable for the best part of a month
469.
by simonw 2mo ago
I wouldn't be too surprised by that, given both the size of the Chinese market and the enormous price discount you get compared to the US models.
470.
by simonw 2mo ago
There's also the private definition reportedly agreed between Microsoft and OpenAI, leaked in December 2024: https://techcrunch.com/2024/12/26/microsoft-and-openai-have-... > The two companies reporte
471.
by simonw 2mo ago
OpenAI didn't have any guardrails in place - they were training a model at a point much earlier than when guardrails start being implemented. The guardrail was meant to be that the agents were running in a locked-down environment with
472.
by simonw 2mo ago
Oh, it's the OpenRouter number: https://finance.yahoo.com/technology/ai/articles/china-ai-mo... Those numbers aren't credible IMO because OpenRouter only see traffic for people who have chosen to ro
473.
by simonw 2mo ago
If someone wants to spend $10/month for exposure to sensationalist AI hype there are a whole lot of newsletters they should subscribe to that will deliver what they want better than I do.
474.
by simonw 2mo ago
The trend I've found most interesting is models of the same size getting better. I'm very much looking forward to seeing how Qwen 3.8 27B compares to Qwen 3.6 27B next week, for example. And the latest DeepSeek v4 Flash has extrem
475.
by simonw 2mo ago
This is part of the training process for a model. They're trying to train it to effectively use existing software to solve problems.
476.
by simonw 2mo ago
> Usage of AI services has shifted dramatically to Chinese providers - from 4% at the beginning of the year to some 30% now. Where did you see that number?
477.
by simonw 2mo ago
I expect I could make a whole lot of money blasting out sensationalist headlines about how the AI labs are all faking security incidents as part of their marketing campaigns.
478.
by simonw 2mo ago
> You're a politician, rather than a truthseeker. Justify that. Also, which "extensive teardown" are you talking about there?
479.
by simonw 2mo ago
Beyond a whole lot of online conspiracy theories I haven't seen anything that suggests to me that OpenAI aren't not telling the truth about what happened here. I find the Black Hat presentation in particular very credible. Also th
480.
by simonw 2mo ago
If you can figure out how to separate instructions from data in LLMs you should ship the first agent system that's guaranteed protected against prompt injection. You'll make millions.

More ›