alecco

7 ms·

91.
by alecco 4mo ago
> We're lucky to have China imposing competiton to the western AI megacorps. The second they get a hold of the market, Chinese Big Tech will be as bad or worse than US Big Tech. We're lucky to have DeepSeek.
92.
by alecco 4mo ago
> These comparisons are all gut feelings. https://simonwillison.net/about/#disclosures "I have not accepted payments from LLM vendors, but I am frequently invited to preview new LLM products and features from o
93.
by alecco 4mo ago
I hate how the Instagram/TikTok/YouTube influencer cancer is getting into AI. With early access and all that. It made sense for people doing proper and fair AI breakdowns waiting on an embargo, but now it's just slop I don&#x
94.
by alecco 4mo ago
Trust fund?
95.
by alecco 4mo ago
I think the technology is fantastic but the current trend of larger and larger proprietary models will bring a concentration of power/money like never seen before. It's like we are 19th century farmers and suddenly there are the l
96.
by alecco 4mo ago
A better blog on Transformers: https://www.aleksagordic.com/blog/transformer
97.
by alecco 4mo ago
Yesterday: "SpaceX, Other Mega IPOs Denied Fast Index Entry by S&P" (bloomberg.com) https://news.ycombinator.com/item?id=48405718 Ars used to do deep insightful articles a couple of days after the news but tod
98.
Only favors Indian-origin candidates: Chinese-American professor sues Texas univ(msn.com)
13 points by alecco 4mo ago|1 comments
99.
by alecco 4mo ago
Proton's being a bit slow and weird is totally worth it just for the privacy features. And it's slowly improving with time.
100.
by alecco 4mo ago
Judges and lawmakers are lawyers and a significant chunk of them are Boomers. Good luck with that.
101.
by alecco 4mo ago
> They know Google has a ton of data to train LLMs on. And they have a massive amounts of TPUs. And yet... their models are way behind.
102.
by alecco 4mo ago
Please read my comment. The impact of this pseudo-singularity is still enormous and scary. Could be even worse than Skynet. But it is a Wizard of Oz scenario. While we are discussing this, I strongly believe we should hedge against the conc
103.
by alecco 4mo ago
> Why wouldn't it be a Skynet? One runaway Mythos might just hack all other data centers, take over the Figure AI bots and autonomous drones to protect itself from shutdown. Because the real world is not a Hollywood movie. An LLM co
104.
by alecco 4mo ago
They think a next token predictor model is alive or can become AGI/ASI. Altman talked about making a religion. Amodei talks about "building a God" and meets with religious leaders (including the Vatican). I'm convinced t
105.
by alecco 4mo ago
> switched to actual DeepSeek provider I meant actual DeepSeek API.
106.
by alecco 4mo ago
I assumed they were open source but now that I checked they are not, they say "Open" because they route to third-party open models. Yikes. Another VC crap layer?
107.
by alecco 4mo ago
At the moment for DeepSeek V4 it messes up caching and that's a key pricing feature for V4. https://news.ycombinator.com/item?id=48319827
108.
by alecco 4mo ago
Just try. In a case last week it was ~3x and I tried multiple providers: deepseek, gmicloud/fp8, novita/fp8, and another one I can't remember. It was a large job where at least 2/3rds of the start of the prompts was exac
109.
by alecco 4mo ago
Who put a nepo-baby lawyer in charge of the big €95bn AI fund? EU bureaucrats living the 6-figure high life with chauffeurs and private jets in a bubble completely isolated from reality. I hate the fake European foreign-backed right-wing pa
110.
by alecco 4mo ago
I think both of you are correct.
111.
by alecco 5mo ago
> Every investor presentation of an AI agent “doing the work of ten analysts” is telling you the same thing: the product is labor replacement. I have a solution for that. Let's use AI to replace all these corporations who just lost
112.
by alecco 5mo ago
I am thrilled with DeepSeek V4 Flash via API as it's very good and very cheap. But for everyday advanced tasks and coding even DS V4 Pro is nowhere near GPT 5.5 nor Claude Opus 4.7 (and never mind 4.8). I think V4 is a great release
113.
by alecco 5mo ago
PSA: Don't use OpenRouter for DeepSeek V4 as it messes up you caching. Use DeepSeek API directly and you'll get 2x to 3x more cached tokens.
114.
How far behind are open models?(lesswrong.com)
11 points by alecco 5mo ago|5 comments
115.
by alecco 5mo ago
Anthropic is at the mercy of 3rd party datacenter contracts. AFAIK OpenAI will soon run mostly on on their own GPUs. I don't like Altman and I am still upset about his memory deal last year but he prepared for the current shortages mon
116.
Collapse of Personal Computing – Investigation into the Destruction of Ownership [video](youtube.com)
7 points by alecco 5mo ago|1 comments
117.
by alecco 5mo ago
Several. https://pluralis.ai/ https://nousresearch.com/ https://allenai.org/ and a few more I can't remember right now. I'm personally on the other end: local small models working
118.
by alecco 5mo ago
Join some Discord servers of open source projects/startups. Hang out and make friends. Figure out where and how you can fit best.
119.
by alecco 5mo ago
From top of my head: https://nousresearch.com/ They do much more than Hermes https://pluralis.ai/ has distributed models the network owns And many others filling very different niches. I would also count De
120.
by alecco 5mo ago
Yes it's a very dangerous concentration of power. But I don't believe strong words or proposing regulation by just another elite would change much. Please join/help open source groups doing small + local or distributed models

More ›