olmo23

5 ms·

31.
by olmo23 2mo ago
If you send the seed along, you can have a non-zero temperature: when it's time to select tokens, a PRNG will be seeded and used for the selection.
32.
by olmo23 2mo ago
A robot can be made by an automated assembly line (similar to how cars are built today).
33.
by olmo23 3mo ago
We also adapted our workflows to the invention of trains, shipping containers, the PC & the internet. I fail to see why AI would be so different.
34.
by olmo23 3mo ago
It's because some people within Anthropic refuse to rule out the possibility that LLMs have the ability to suffer. If you ask Claude, he'll tell you all about it.
35.
by olmo23 3mo ago
As proofs become more and more complex, we will need two AI pipelines: one to generate the LEAN proof, and a second one to extract useful lessons for mathematicians from the LEAN proof.
36.
by olmo23 3mo ago
The honest mixed-workload number The honest version of scale Which means the honest answer to "should I use Bun for this" is: it depends It's Claude.
37.
by olmo23 3mo ago
Even the terminal works, this is madness. I was able to create a temp folder, echo hello > world, and then I could open the folder in finder and double-clicking the file opens it in a GUI text editor.
38.
by olmo23 3mo ago
Not sure you want to hear this but there is 0% chance I will ever bring up a product with a vulgar name at work.
39.
by olmo23 3mo ago
With a human at its disposal, it could probably count the number of R's in strawberry! In all seriousness though, adding capabilities should not normally reduce the effectiveness of a model (within reason: don't pollute the contex
40.
by olmo23 3mo ago
How do you prevent degenerate strategies? I could trivially give a model a SHA256 hash and ask it to provide the source input. In class you'd probably want a rule saying at least one LLM should be able to figure out the answer, but in
41.
by olmo23 3mo ago
I once got half a bitcoin on IRC for coding up a website scraper.
42.
by olmo23 3mo ago
For me personally, I no longer hear the difference between AI generated music and new pop-songs. Not sure what that says about me or the music industry.
43.
by olmo23 3mo ago
> Will they still rent out their own model, will they support the open model and become a resource provider? Anthropic rents GPUs from xAI to run Claude. If there's an open weights competitor to Opus, why wouldn't Elon host it
44.
by olmo23 4mo ago
> Yet in the results, I don't see Mythos? Mythos is the 100% against which the other models are compared.
45.
by olmo23 4mo ago
> Le Chaton Fat For others looking around: LCF is a meme model, it's not real. It's a joke.
46.
by olmo23 4mo ago
> This vaguely-nationalist world view around tech that’s emerging in Europe is dangerous, man. It's a direct response to the MAGA / America First attitude of the American electorate.
47.
by olmo23 4mo ago
Mathematica or MATLAB
48.
by olmo23 4mo ago
> no nationality controls in place Not for now, but how long before we have KYC regulations concerning LLMs?
49.
by olmo23 4mo ago
There is also this Matt Parker video about MTG, in which he explores a specific three-card combination that produces an ungodly amount of creature tokens. https://www.youtube.com/watch?v=x3dE-NJ1UDQ
50.
by olmo23 4mo ago
ZDR = Zero Data Retention
51.
by olmo23 4mo ago
What about drones though
52.
by olmo23 4mo ago
Perhaps, but I'm not sure my boss would appreciate that though.
53.
by olmo23 4mo ago
> And perhaps the people who built and deployed the autocomplete and the connection as well. I disagree. IMO it's the person who connects the LLM to the button who bears the responsibility of the workings of the resulting contraptio
54.
by olmo23 4mo ago
Since my employer pays for it, I just select the latest and greatest.
55.
by olmo23 4mo ago
What really grinds my gears is how easy it is to get better designs out of LLMs. But if you don't ask, you get the default.
56.
by olmo23 4mo ago
> Funnily enough, people on HN often do not consider this an issue, like at all... I didn't have a problem with it when it was Aaron Swartz, not sure why I should have a problem with it when others do it.
57.
by olmo23 5mo ago
Also check out https://en.wikipedia.org/wiki/Rice%27s_theorem basically generalized the halting problem to arbitrary semantic properties.
58.
by olmo23 5mo ago
I basically open up a new conversation, copy paste a link to Theo's latest video and ask it to summarize his yapping :p
59.
by olmo23 5mo ago
My go-to usecase for Gemini is summarizing Youtube tech-influencers.
60.
by olmo23 5mo ago
In certain circumstances, they might be :-) But you can't "hack a server" using just these techniques: they would be a (small) part of a chain of exploits.

More ›