visarga

6 ms·

31.
by visarga 8d ago
Author might have spend days researching it, selecting what to include and what not to, revising the language and graphics, then comes Pangram and slaps 100% AI on it without looking into the substance.
32.
by visarga 9d ago
When AI gets that good, a solar panel can power a pile of motors and gears to move and do work for you that previously required humans. You won't need the job, you need to solve your necessities more directly. And that job is a good fi
33.
by visarga 9d ago
Haven't you seen how ridiculously good robots have become?
34.
by visarga 9d ago
And because of you we are now looking bad, so you're pressuring us with your unleashed creativity.
35.
by visarga 9d ago
Maybe they still want to serve the feee tokens if that gets them good training data.
36.
by visarga 9d ago
On macs you just install a desktop app. What can be simpler? Decent Qwen models already doing some simpler types of agentic work.
37.
by visarga 9d ago
They can't afford to miss out or be seen as missing out. It's not rational, yet it is in a way.
38.
by visarga 9d ago
Yeah, I have a room like that but I never stay in it. No screens.
39.
by visarga 9d ago
Going on social media to talk about not going on social media. It's at least 50% as effective as actually not going on social media.
40.
by visarga 9d ago
7.5 years ago it was perfect
41.
by visarga 9d ago
>> I think the „my brain isn’t the same anymore“ sensation is pretty universal. Another pretty universal thing is that we compare to younger versions of ourselves. That might explain the gap.
42.
by visarga 9d ago
This only holds when you keep the task fixed and compare with and without AI. But because AI changed expectations the difficulty of work we supervise today is much higher. My theory is that AI runs our brains just as hard as no-AI, maybe ha
43.
by visarga 10d ago
I've been RSI'ing for 6 months.
44.
by visarga 10d ago
> That company is morally bankrupt When they fired Sam 700 out of 770 OAI employees threatened to move to Microsoft together. So they were giving their work on AGI to MS just like that.
45.
by visarga 10d ago
They just released Astra, claimed it is AGI. The slowdown begins immediately after OpenAI's jump.
46.
by visarga 10d ago
Some people type 100x more words than the final article just digging into a topic and turning it on all sides. The final article being generated tells you nothing of the size of the effort going in.
47.
by visarga 12d ago
It's like finding random hornet nests.
48.
by visarga 13d ago
> can match or exceed human cognitive abilities across a wide range of tasks If you go by definition AGI is not general, just "smart ape" shaped.
49.
by visarga 13d ago
It's not measuring AGI at all, it starts from human "core knowledge" so it is parochial. It is made of tests that still fail so by definition next version will also start low. Moving goalpost.
50.
by visarga 13d ago
Their definition of AGI is "when we can't invent any more tests where it fails"
51.
by visarga 13d ago
I'm doing the opposite - prose. And specifically for Claude the 4s - "simple short synthetic style"
52.
by visarga 13d ago
I care more about the recursive loop between execution and cost. Anything that executes - be it biological or artificial - incurs some costs. Those costs in turn impact what execution can follow. The gains pay for action, action produces ga
53.
by visarga 14d ago
it's called `agy` now
54.
by visarga 14d ago
I am often asking it to write in sss-style - "synthetic, short and simple style"
55.
by visarga 14d ago
It's so bad I made my own chat client for Claude, so I can attach steering prompts in conversation. They are applied just at the end, before the last LLM response, then removed and response kept.
56.
by visarga 14d ago
> when you increase thinking-effort, that seemingly increases response-length I use the /sss writing style - synthetic, short and simple - and it helps a lot.
57.
by visarga 14d ago
You can engineer loops that have it, but it depends on a case by case basis. Does your loop have strong validation? if it's all vibes nothing can stop it from diverging.
58.
by visarga 14d ago
> Good software will be good-er. Bad software will be nightmare fuel. We are talking about a moving target here ... they get better every few months, so I expect the super-LLMs from 2035 will write amazing code even with sloppy prompting
59.
by visarga 14d ago
> Prompt Processing (PP) 66.3 tok/s I got 400 pp tps on a 10k token input. Your numbers seem suspiciously low, maybe the input was too short to measure properly? And this dense 27B is slow, the MoE A3B models get to 1000 tps.
60.
by visarga 14d ago
Local is never cheaper than cloud because they can do batch inference, and that means you load model weights once to produce 128 tokens on 128 sessions in parallel not 1 token on 1 session like local models. Local models rarely get to high

More ›