rtrgrd

Could we add (2016) to be title?

Reading charitably: just because our inventions do something different to nature doesn't mean that our invention is wrong;

In context: just because our LLMs don't have an explicit 'system 2' component doesnt mean it can't have superhuman reasoning

Oh man all of his Chinese humour is lost in translation T_T

Just wanted to say a huge thankyou for being so patient in the forum; it's quite annoying that the comment section is a more a function of the title + personal opinions than a function of the blog content.

I love using obsidian, and thanks so much for all the work that you and the team have put in :)

Confused here - setting up certs to MITM https requests to add a header seems like a decently big security risk?

Also similar to block withholding [1]

[1] https://bitcoin.stackexchange.com/questions/4943/what-is-a-b...

I also see 577, so must be human testers action count. Also watched the replay, the solve seems different to mine.

Added to my non-llm username list :)

Thanks so much for the explanation

Can we please flagban

Hyrums law at its finest :D (or D: if you deeply care about correctness)

Very cool - note that lowercase b, l and h are the same

I would suspect that for self hosted LLMs, quality >>> performance, so the newer releases will always expand to fill capacity of available hardware even when efficiency is improved.

All the hedge funds sniping orders right now lol

Might be hug of death but the load times are horrifically slow.

The blog mentions checking each agent action (say the agent was planning to send a malicious http request) against the user prompt for coherence; the attack vector exists but it should make the trivial versions of instruction injection harder

we all love non ascii code (cough emoji variable names)

I thought human preferences was typically considered a noisy reward signal

I've never used Kagi before and wanted to try: how does Kagi stack up against Brave search?

I assume the high volume of search traffic forces Google to use a low quality model for AI overviews. Frontier Google models (e.g. Gemini 2.5 pro) are on-par, if not 'better', than leading models from other companies.

Generally agree with the ideas of the article and wanted to know more about what a 'makerist' movement for AI would be like.

Quite concerning to see the issue still marked as open (since jan!), hopefully it got fixed and it's just that no one marked as closed

Helium can be produced by alpha decay (hence unlikely to run out in the near future) and is also obtainable from natural gas. That being said it is still non-renewable and has quite volatile prices for some reason.

I know it says that it's a for fun project in the description, but something about normalizing sending passwords in plaintext via HTTP to a non-deterministic security checker sounds like a bad idea ;) Still a funny project tho

afaik if you set up your own headscale server, i don't think there is a device count limit (?)

Thanks for picking that up - fat fingers

Am I the only one who is experiencing severe lag when scrolling on the new site (Firefox android)?