villish

8 ms·

1.
by villish 4d ago
Add in a heaping dash of anti-american sentiment, and you will get the truth behind the pessimistic commentary. Downplaying the latest models capabilities is frankly insane considering what we’ve seen what OpenAI’s models have done without
2.
by villish 4d ago
> Qwen 3.8 and GLM 5.3 Flash were not just cheaper, the results were significantly higher quality How can they be higher quality than models they were distilled from?
3.
by villish 5d ago
Unusual means? It’s illegal to steal accounts and credit cards. Completely sanctioned by PRC. No other country can compete since they must follow the law, hence mistral not even trying anymore.
4.
by villish 5d ago
> these proxy services create thousands of new accounts using false identities, fake or stolen credit cards, and stolen API keys. They will often use stolen API credentials belonging to legitimate companies or individuals to give unautho
5.
by villish 5d ago
> Framing learning from observation as somehow bad. How about hacking accounts, using stolen credit cards, and sending data to the US silently? > Pretending that this is bad It is. This makes AI a 2 horse race because European labs ca
6.
by villish 5d ago
Except they use thousands of stolen credentials which is definitely not legal. They also don’t notify users they send data to US labs which can include sensitive information. EU labs like mistral cannot legally do any of these things. So pe
7.
by villish 6d ago
They mean distillation is legitimate when labs use one of their own stronger models to train a smaller one. They certainly aren’t advocating for PRC labs to distill Claude for open weight models.
8.
by villish 6d ago
The list of professions made obsolete is very long. Why should programming be protected if AI in the future can make software better, safer, and cheaper?
9.
by villish 7d ago
Yes, I have Tailscale on my AppleTV. Only traffic attempting to reach your tailnet is routed to you, the rest is on his own network. If you want to make it super easy for non technical people also use reverse proxies to your tailnet so jell
10.
by villish 8d ago
If they could harness the energy from all your atoms it would be worth significantly more than 2 million.
11.
by villish 8d ago
I hate it because I hate having to look at these AI generated images. In fact I want an internet where I have to explicitly opt-in to see AI generated content in general.
12.
by villish 8d ago
My apologies then. Those debrid services seem like they are run in on borrowed time though so I don’t use it. I have friends that do and seem to like it, but I prefer having a local library (*arr stack)
13.
by villish 8d ago
Tailscale. Your friend just needs to be invited to your tailnet. Super easy to use with clients on all the major devices.
14.
by villish 9d ago
This seems like spam in order to get people to sign up to a service with your referral link. It’s literally just a guide on how to create an account.
15.
by villish 9d ago
I like them both, but I can never give up plexamp. I’ve tried navidrome and various jellyfin clients but I keep going back.
16.
by villish 11d ago
(58% on OpenRouter) I believe both things can be true. OAI/Anthropic selling more tokens than ever, and downloadable weight models increasing in market share. I doubt OpenAI would copy Fables pricing if it wasn’t financially beneficial
17.
by villish 11d ago
> different benchmarks for different domains. Labs know the only thing the public even discusses on model releases are benchmarks, so they devote a majority of training on just benchmaxxing. It’s marketing. Muse Spark looks great in benc
18.
by villish 11d ago
That likely won’t change if other competitors don’t take the lead at some point. If companies are willing to pay top dollar for the best models AND they get to extract as much money from Chinese labs distilling Astra/Fable it makes no
19.
by villish 12d ago
Hopefully this wakes people up from this addiction to benchmarks when discussing various AI models. Different model families have strengths and weaknesses in various domains, but those are never discussed.
20.
by villish 12d ago
“Bobcoins are the consumption-based billing metric used for Bob subscriptions. They provide a transparent, predictable way to measure and bill for your Bob usage.” I’m getting sick of these mobile game currency style credits/coins that
21.
by villish 13d ago
I have no faith in these benchmarks. Muse 1.3 shouldn’t even be in the same conversation yet it scores above GPT 5.6 & 6.0
22.
by villish 13d ago
Frontier. Everything is frontier. K2 not to be confused with the other K2, or K3 that is also frontier.
23.
by villish 13d ago
Yeah and how was it done? Why did all (60?) of the trump campaigns lawsuits get thrown out due to lack of evidence? Why did the democrats just simply forget to do it again? How many hoops will you have to jump through to answer these questi
24.
by villish 13d ago
Pretty much. Europe is basically not at all involved in the AI race and is reliant on US and China.
25.
by villish 14d ago
Explosive/Chemical based is my line of thinking. A Fable level model with zero safe guards planning and coaching someone is within the realm of possibility now. That is much more dangerous than a man with a gun.
26.
by villish 14d ago
To have at least a choice of using a European trained model. Downloadable weights is not open source. Mistral is able to respond to any regulatory queries about training data and concerns.
27.
by villish 14d ago
> Or is this just a dog whistle for the HN crowd who like to complain about anything related to EU regulations? Only topped by the number of people who love to cheer for US vs China while living in neither.
28.
by villish 14d ago
> These are not really analogous: direct lethality of LLMs is quite low. I imagine an LLM with no safeguards and a psychopathic mind would be considerably more dangerous than a gun. I don’t mean only for hacking. Though everyone in the
29.
by villish 14d ago
Should everyone have access to guns too? I ask because there seems to be a disconnect where a lot of people who live in countries with gun control don’t want AI offensive capabilities to be regulated. I presume your country doesn’t have AI
30.
by villish 15d ago
Keep in mind these downloadable models use 3-10x the amount of tokens as well. You really can’t beat a couple $20 subscriptions. https://quesma.com/benchmarks/babaisbench/

More ›