HNHacker News
TopNewBestAskShowJobs

notrealyme123

526 karma · joined April 10, 2021

submissionscomments
notrealyme123··on GitHub Down Again?
And if you put it in private we will never know, but pinky promise. They would never steal training data
notrealyme123··on ISponsorBlockTV: SponsorBlock client for all YouTube TV clients
Would love to look at it, but GitHub gives 504.
notrealyme123··on GPT-6 Astra covertly underperforms on safety research
A few weeks ago that companies agent hacked another company to solve a task by whatever means possible. Where exactly are they overfitting for Safety here?
notrealyme123··on Progress on provable inference for large language models
Claude being a official contributor here gives me the shivers. This feels like setting up a case for: We have the rights for everything you do with Claude. Or rather what they do with Claude for you.
notrealyme123··on Progress on provable inference for large language models
It would be fantastic to compare benchmark values. The current trust-me-bro economics leave little other choice than interpreting those numbers as PR.
notrealyme123··on Progress on provable inference for large language models
I really don't like that the first more elaborated real-world case is to stop people from training their own models. (Challenge 1)
notrealyme123··on Three new discoveries in mathematics, computer science, and physics
I guess I am not doing my PhD in an actual research lab then. We don't get Claude/chatgpt paid, but we basically have to use them. Otherwise we don't have any chance for sota research. And I am not talking about AI doing innovation. Just shorten the time between idea and results.

Right now we have to Pay it ourself, and we are very aware that oai might use our research.

notrealyme123··on Root shell on Samsung TV
There is more written here: https://www.kimi.ai/blog/kimi-k3-1-pro

They claim that the default datasets contain so much cyber stuff, so that their models became quite good at it.

They filtered the data before training to not contain this. So it's not that they train the model to not do it. They rather not have it in the Corpus at all. In theory

notrealyme123··on Root shell on Samsung TV
It creates a lot of uncertainty. Especially if you depend more and more on them in your day to day life.

If your always-on agent hears to wrong thing, you could lose access to things you depend on. The same reason makes GCP very unattractive.

notrealyme123··on Hacker reverses Samsung TV, finds on-board AI listening OpenAI account suspended
Iirc They specifically trained them to not be useful in cyber stuff. But the point still holds
notrealyme123··on Today, we hacked what we'd set out to hack: the iPhone
Correct me if I am wrong, but you don't get root access via the android developer options.
notrealyme123··on DeepInfer: High-performance inference framework for DeepSeek models
20 merge requests in the last two hours. And all look somewhat reasonable on first glance.

I am happy for not maintaining a relevant open-source repo.

notrealyme123··on AOSP Source Releases Are So Late Developers Disassemble Libraries to Report Bugs
Interesting theory. Is there a distinction between phone release and os release there? I would imagine the time starts when the model goes in sale, not when the os releases.

Otherwise I would expect Google would also wait with pixel updates

notrealyme123··on DishwasherBench – Can Frontier Models Stack a Dishwasher?
I think putting cups with the opening facing upward is very close to "technically correct, but very wrong"
notrealyme123··on Imposing Sanctions on the International Criminal Court
> "Either the ICC will end its threats, or we will end the ICC."

Are the ICC already terrorists, or not yet? Asking for a friend who doesn't want to get bombed.

notrealyme123··on OpenAI, the Partition Principle, and Mathematics
Like all the junior devs getting opportunities left and right.
notrealyme123··on OpenAI, the Partition Principle, and Mathematics
If you can't leave them, love them
notrealyme123··on Navier–Stokes Lost in Translation
I get the feeling a lot of people propose that we can write a verifier for every proof in lean.

Can someone tell me in simple terms why this doesn't conflict with the incompleteness theorems?

edit: thanks for the responses, i feel slightly less dumb now

notrealyme123··on Rejection Sensitivity in Gifted and Twice-Exceptional Children
How do you use autism to filter something? Did you train this?
notrealyme123··on Nvidia wants to put a watchdog chip next to every AI agent
The discussion right now brings this framework into the training corpus of the model.

Give it a year and it is outplayed by new agents. Shovel makes have to sell shovel's

notrealyme123··on Claude is only available to people over 18 years
This data does not support the argument made here.
notrealyme123··on Jellyfin 12.0
I am currently on vacation and used jellyfin via the browser on all the smart TV. And it just works great. I didn't expect it to work this good.

Sign in via key is easy, not stuttering. I can recommend it.

notrealyme123··on Claude Fable 5.1 and Claude Mythos 5.1
No idea if subagents are not blocked, but you can set the starting method for subagents to "fork", then they inherit the main agents context
notrealyme123··on Ox Alpha
On matched hardware?
notrealyme123··on Is AI reasoning right for the wrong reasons?
The assumption is that those reasoning traces are the lowest level. But we can't observe the reasoning-for-reasoning (the reasoning in a transformer block.)

Saying that one is enough is pretty arbitrary.

notrealyme123··on Codeberg Bans Cryptocurrency Projects
So I am very pedantic here, but yes it "can" be seen as political. You made the decision that your free time is more important than working and generating value.

I understand "everything is political" more in a sense that we as a society have the burden to shape our own rules, and saying "I only live in the society, but ignore the burden of making rules" is not really possible. In that case you actively decide to let someone else decide for you.

notrealyme123··on The EU is about to sell our most sensitive data to the US for visa-free travel
It's ok to be angry, but this is just wrong.
notrealyme123··on Germany maybe found a new source of renewable energy
no brain in the truest sense of it.
notrealyme123··on Patching MechCommander's "left arm bug" for fun and profit
A bit off topic, but a very nice niche YouTubers "SardonicSays" posted a great Retrospective about MechCommander yesterday.

https://www.youtube.com/watch?v=8CVvOWlfl1U

notrealyme123··on US Food and Drug Administration rejects petition to set PFAS limits in food
I can not find it in the FDA list. Is there a newer source?
Page 1 of 8Next →