krapht
krapht on Hacker News
Should mention Jax in title, otherwise people won't know why they should use this instead of torch.
spinning up a venv with duckdb: 5 minutes
using sql everywhere; never having to remember dataframe syntax: timeless
Yep. Nobody has repealed Medicare, for example
His example of a web crawler stinks. I can get behind the philosophy, but his specific web crawler example at a minimum needs to retry, respect robots.txt, and rate limit per domain. So no, xargs + curl is a dumb example.
Which is why I've never quite understood the appeal of portable SIMD libraries for performance-critical code. If I'm explicitly writing SIMD rather than relying on the auto-vectorizer, it's usually because I want access to the particular capabilities of the target ISA.
For many problems, choosing the right instruction or instruction sequence makes a large difference. Portable SIMD abstractions necessarily expose some common semantic layer, but SIMD ISAs don't actually have equivalent capabilities. Instructions like pshufb, for example, enable algorithmic tricks that don't necessarily have an equally efficient analogue on another architecture.
If maximum performance matters, I generally want intrinsics and architecture-specific implementations; if portability matters more, I'd rather move further up the abstraction stack and use something designed to target multiple architectures, such as ISPC. There are certainly cases where portable SIMD gets close enough to optimal, but I don't think there's a compiler or abstraction that can express every useful SIMD idiom and lower it equally efficiently across fundamentally different ISAs.
One I spent a few months working on was pathfinding for trucks. The goal is to find dominant solutions over {shortest time, lowest cost (tolls + fuel), avg road speed variance - traffic sensitivity} and then return 3-4 routes that are equal distance from each other in this dimensional space for users to pick from.
As you say, the most useful things happen in low-dimensional spaces.
social media is a completely different market though - since there are massive returns to scale, it's incredibly hard for a new entrant to break in.
model training and inference is the opposite. people switch LLMs like people change clothes in the morning. there are popular services (openrouter) that make moving as easy as changing a model string.
Is Gemini really doomed? I'm still bullish on Google:
1) they have more free cash flow and capital than God due to the ads business
2) they have data - intent from web searches, youtube videos, google books and music
3) they have dedicated inference hardware
for all these reasons, is being 6 months behind the frontier actually a structural, long term disadvantage? some day the pace of improvement will slow, and google will vacuum up the market. they'll be able to compete with open-weight models just on pure cost advantage from their vertical integration
I think that's backwards. Anything that changes the performance of a model deserves a minor version bump. A new model has to be qualified before being pushed to production; but we don't get the choice here, just cross your fingers there are no regressions at all on all possible tasks the model might be asked to do.
News to me. What do you call mixture-of-experts and model distillation?
ironically, any frontier LLM will easily generate a tutorial at any detail you like explaining what these are.
if don't have time for that, just know that these are technical parameters that affect how likely it is an llm will produce the same result after being asked the same question.
It's more rare, and definitely less lucrative, but there are opportunities in software engineering for non-profits, open-source, and smaller non-tech companies. If you've made your blood money and don't need top n% pay, it's something worth checking out.
Slight tangent: a long time ago when I was into this hobby, Oshpark was usually the go-to instead of JLCPCB. What happened to them? Macrofab too if anyone's used them.
> The catch is that for image resolutions >=700x700pixels (most production usecases)
Citation needed? 2XL looks like you go up to 800x800 pixel inputs. This isn't the dealbreaker you say it is - all pipelines benefit from thoughtful crop and rescaling before going to inference.
More than just your commute. Any downtime where you are doing a mindless chore is prime podcast time.
I've heard of people using anythingllm for this purpose.
Basic rag is almost stupid in how easy it is, though. You grep for keywords, take the surrounding paragraph, then stuff it all into your llm prompt.
The next upgrade is to automate keyword extraction by putting your documents into a vector store and search by vector similarity.
I went and looked at the code. It's AI bots and a few confused trend followers all the way down. There's no way anything in there works.
<article excerpt>
What AI actually does
Mutinex has built what it describes as a “multi-agent system,” where each agent acts as a domain specialist. For example, one agent understands marketing econometrics, another understands competitive pricing theory, another diagnoses model failures.
By combining Tracer, which cleans and makes sense of Hershey’s data infrastructure, with Mutinex’s AI system, Hershey is now able run models in as little as three weeks.
In practice, that means faster iteration on how marketing spend is evaluated and adjusted, rather than waiting for lagging historical reads.
“Most companies don’t have an AI problem. They have a data readiness problem,” said Sarah Martinez, chief commercial officer, Tracer.
</article excerpt>
Instead of the headline, it sounds like they've hired an external company to clean up their ETL pipelines. That seems useful.
I'm going to doubt spooling up <massive LLM> with <appropriate system prompt> is going to be the thing that reduces their analysis time.
You can go your entire career without recursing, or using a tree data structure in its raw form (i.e. you only use it as part of a library)
which city is that? I want to move there with my children.
If you desperately want mediocre chromebook instruction for your children; where your classmates turn in ChatGPT essays as original work, feel free to move to Virginia.
Lots of sensors these days will give you 10 or 12 bits of data per color channel. You may want ordered dithering when previewing on an 8 bit display.
Is the code actually good though? Not seeing any benchmarks vs numexpr, numba, or Jax
Sublime Text? Sure, doesn't have the long tail of extensions, but surely most people don't need those. The biggest issue with ST being the fact that it costs money...
Paying customers use docker desktop.
Even if you don't like the current administration, the rank and file are still out there doing valuable work. The government is more than ICE; it also administers welfare, funds research, collects taxes, and distributes social security payments to the old and infirm.
So? If the goal was to attract corporations you'd release under a BSD license.
Why? Is Scotland innocent?
This post makes me sad for two reasons:
1) kids k-5 are using laptops
2) websites are so bloated today that browsing on a Chromebook is causing agitation
Otherwise I don't have anything topical for this post, but good luck OP
Yes, but is this a problem? Haven't most betting markets turned out to offer accurate predictions?