sudb

5 ms·

31.
by sudb 3mo ago
> but it ain't there On what grounds? I don't think you've provided any evidence other than LLMs can't "adapt" or "learn" to show that LLMs do not show intelligence in any way. I think it's
32.
by sudb 3mo ago
> To say that LLMs' existence is evidence for emergent phenomena in LLMs is tautological. This is not at all what I was saying. I think you've already conceded that LLMs demonstrate emergent behaviour but you dismissed it as a
33.
by sudb 3mo ago
> Figure out how to make it happen reliably And you're happy that the replication of LLMs across many foundation model companies is insufficiently reliable? > just something you're imagining So the alternative explanation yo
34.
by sudb 3mo ago
Ah it seems you are a stochastic parrot believer ( https://en.wikipedia.org/wiki/Stochastic_parrot ). What are your responses to the Expert Rebuttals section? I find the rebuttals pretty convincing - that there seems to
35.
by sudb 3mo ago
> By reaching into the voids of its embedding space and returning tokens related to nonexistent semantics. Or, if you like, "hallucinating". The hallucinations which are useful we might call "novel". This sounds to me
36.
by sudb 3mo ago
I think a few years ago (perhaps it's still ongoing) there was a big uptick in purchases of "old" compact digital cameras - I still see young adults and teenagers taking pictures of these around London and I think this is a v
37.
by sudb 3mo ago
> Google "temperature in LLMs". No need to condescend, I'm very aware of what temperature is for LLMs. But I'm going to push back - if you're claiming all LLMs simply do is a stochastic _search_, how can that pro
38.
by sudb 3mo ago
Is an implication of this that models are incapable of producing entirely novel code? Also, not to get too reductionist about this, but what do you posit is special about what is happening when humans think? Intelligence is hard to define s
39.
by sudb 3mo ago
I found this article very well written, as a comparison https://explained-from-first-principles.com/internet/
40.
by sudb 3mo ago
This feels like a semantic disagreement to me? If an LLM got to an acceptable end result code-wise, what would you call the process that took place to get it there?
41.
by sudb 3mo ago
I am surprised that no comment has yet mentioned Infinite Jest (which I'm still yet to complete - it's been years since I started).
42.
by sudb 3mo ago
Sounds very reasonable - I don't have a robot vac at the moment because there are some very large thresholds in my flat and the floor itself in places is very wobbly, but if I had a guide to safely mod one to work for my residence I&#x
43.
by sudb 3mo ago
Also, in my opinion, it's much easier to build a good MCP interface than it is to build a good CLI interface - and afaik there's support for MCP tools to return things like images from an MCP tool call directly to the calling LLM
44.
by sudb 3mo ago
For IDEs there are many non-Electron apps out there. I like Zed, for example, and no longer have any other IDEs installed (not counting Xcode, or emacs/vim if you're one of those).
45.
Generating Images with a 2025 Android(duration.ai)
28 points by sudb 3mo ago|0 comments
46.
by sudb 3mo ago
I've managed to successfully use the ANE to accelerate text-to-speech models on iOS (as an aside - this was much more straightforward than the equivalent on Android). I did however struggle to run a diffusion model on the ANE - but fou
47.
Denoising Voice Recordings On-Device(duration.ai)
1 points by sudb 3mo ago|1 comments
48.
by sudb 3mo ago
I applied some reflective film to the inside of some of my windows today and it seems to have had some effect, though the window is pretty hot to the touch now and apparently this kind of film can cause windows to crack through thermal stre
49.
Generating Images with a 2020 iPhone(duration.ai)
3 points by sudb 4mo ago|0 comments
50.
by sudb 4mo ago
This is the "company brain" product spike from Anthropic I think https://www.ycombinator.com/rfs#company-brain
51.
by sudb 4mo ago
There was an interesting study that I attempted to enroll in (but unfortunately was screened out by both age and childhood instrument learning) that was looking at psilocybin potentially giving adults the ability to acquire perfect pitch:
52.
by sudb 4mo ago
Multiple successive very short sentences are also anecdotally an LLM tell I think
53.
by sudb 4mo ago
Something elided here is that nested virtualization on regular EC2 instances has only been possible since February this year[1] - before this, you had to use a metal EC2 instance to run Firecracker VMs. 1. https://aws.amazon.co
54.
by sudb 4mo ago
Very interested to see where this kind of work goes for on-device video generation!
55.
by sudb 5mo ago
If this means that segfaults become rarer with Bun I might consider using it in production again. As it stands, Bun has been great as an all-in-one TS/JS package manager, build system and test runner but unstable enough that I still wa
56.
by sudb 5mo ago
sure, but I was replying to a complaint about failure to parse, not existence of parsing isn't the real pragmatic answer to parse out known invalid email address strings, and only attempt to deliver an email if unsure?
57.
by sudb 5mo ago
> the TAM is exploding at a step function rate This stumped me for a while! what are the interpretations, infinite or 0?
58.
by sudb 5mo ago
Famously valid email address parsing is far from trivial[1] - I wouldn't be so quick to judge! 1. https://www.regular-expressions.info/email.html
59.
by sudb 5mo ago
I think a lot of time and energy perhaps gets mis-spent on this but the title is right - if you can watch it, you could use any one of a number of methods (incl. capture cards) to record whatever is on your screen, and to some extent I mayb
60.
by sudb 5mo ago
This is first time I have experienced some vindication that choosing CUID2[1] for one of my projects was actually a good idea. 1. https://github.com/paralleldrive/cuid2

More ›