Hacker Newsnew | past | comments | ask | show | jobs | submit | sandos's commentslogin

Ive ran GPT-6 and 6.1 models repeatedly through multiple-hour sessions, they dont stop that often.

This explains why I like GPT: I LIKE for them to excessively test each step, kthx!

I doubt "any 4V power source" is true AT ALL!

I once tried powering an android tablet with straight DC, turns it is EXTREMELY sensitive to slow power supplies, ie. it required a very beefy one. I had to add capacitors even to get it to work, it was really weird. But considering a battery can dish out some serious amps....

Also 4V sources are not common.


Received 3 emails in the last 2 hours. But widely communicated that "Teams is the way now!".

Luckily I can sometimes forget to check my email for an entire day, and it nobody even notices. At least _I_ am not that dependent on it.

Also some safe browsing thing is offline.. we just click "disable it and continue"... people are feeling very safe indeed.


My BS-meter just went through the roof:

"Specially configured AMD hardware, which is also restricted from export to China, is the only alternative we have identified; we have excluded it for DeepSeek-V4-Flash"

This sounds as if they are using something VERY fuzzy to "match" things, and if they are ALSO matching against completely different hardware which this makes it sounds like.... then I dont know if this means anything.


They match the pseudo-random generator from the NVIDIA library cuRAND.

AMD has a library that emulates it, for compatibility.

Their methodology is interesting, but it only proves access to a single Blackwell GPU, which would have been sufficient for initializing an LLM with random values, before the start of its training.

There is no evidence whatsoever about which GPUs were used for training, except the supposition that if they had access to 1 Blackwell, they might have had access to more of them. Such a supposition may be justified about the number of cockroaches in a kitchen, but less so about the components of a training cluster, which might be heterogeneous.

The LLM could have been initialized on the workstation of a developer, before being transferred to a server cluster, for training.


Sweden is ofc not very dependent on gas at all, but we are indirectly coupled via the grid to gas prices, this is well known.

I would be happy with 5 unlimited agents. We are limited by models and the amount of credits (copilot) we get, so...

I can't even imagine what I would so much compute for. My private 10$ copilot plan was basically unused last month, I could just not come up with anything to spend it on. (Also suffering a bit of AI burnout from work helps here...)


I feel strongly that GPT-6 Luna has been nerfed. I was super-impressed by it in the beginning, now it seems to think overly much and get absolutely nowhere with even very simple tasks.

Wikitree is also fascinating, to me at least! :)

What do people mean that VS code can not support 12 sessions? I never do that, but 3-4 in parallel is fine, and I don't see why it would not support many more.

I do use multiple windows though, I never liked mutiple workspaces in the same window, it also makes contexts blow up I find.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: