j2kun

12 ms·

241.
by j2kun 1y ago
I think this hits the nail on the head: performance is the obstacle, and you can't get good performance without some modeling expertise, which most people don't have.
242.
by j2kun 1y ago
This is what major sports leagues use for season scheduling (source: https://mathstodon.xyz/@j2kun/108975072813565989 )
243.
by j2kun 1y ago
> The real advantage of solvers, though, is how well they handle new constraints. Well said. One of the big benefits of general constraint solvers is their adaptability to requirements changes. Something I learned well when doing datacen
244.
by j2kun 1y ago
If you've ever lived in a place with a localized car ban, it's awesome.
245.
by j2kun 1y ago
Build systems are yet another special circle of hell.
246.
by j2kun 1y ago
> The way to find the right answer isn't to think very hard and miraculously come up with the perfect answer. It's to try every goddamn thing that shows promise. I have found that spending more time thinking generally reduces t
247.
by j2kun 1y ago
What are you ranting about? Lattner has a strong track record of producing valuable, open source software artifacts (LLVM, Swift, MLIR) used across the industry.
248.
by j2kun 1y ago
Yeah the OP here listed a bunch of Python stuff that all ends up shelling out to C++. C++ is rich enough, period, but people find it unpleasant to work in (which I agree with). It's not about "richness," it's about givin
249.
by j2kun 1y ago
https://arblib.org/#special-functions has all manner of distributions available. I think you misunderstand what I meant.
250.
by j2kun 1y ago
The frequency with which people involved in cryptocurrency "pull the levers themselves" has far outpaced government manipulation of currency.
251.
by j2kun 1y ago
And people could just do all their business in cash to avoid banks. But that's not practical just like avoiding exchanges and not using wallet providers is impractical. Normal people cannot function in a cryptocurrency ecosystem withou
252.
by j2kun 1y ago
In this case, Stripe is adding themselves as a middleman. Whether or not it was the point of Bitcoin from the start, "removing the middlemen" is bullshit because you still need exchanges, wallet providers, people running nodes, et
253.
by j2kun 1y ago
> Over the last few years, stablecoins have become a preferred means to hold and move money (for convenience, etc). For avoiding regulation.
254.
by j2kun 1y ago
All of the other comments are missing the point: using blockchain technology is a means to bypass regulation. That's it. That's always been the point of cryptocurrency.
255.
by j2kun 1y ago
> more powerful than git > not missing out on power Two very different claims, and it only makes me more skeptical.
256.
by j2kun 1y ago
15 years of experience and you're a few months from being homeless? My friend, there's a lot you're not telling us.
257.
by j2kun 1y ago
This concept has been done many times in the past, under the name "interval arithmetic." Boost has it [1] as does flint [2] What is really curious is why, after being reinvented so many times, it is not more mainstream. I would lo
258.
by j2kun 1y ago
Summoning Hillel Wayne.... https://www.hillelwayne.com/post/are-we-really-engineers/
259.
by j2kun 1y ago
Yeah I can't find any example in an AGENTS.md that isn't useful information for a human. "Cluttering" a README.md is a silly justification.
260.
by j2kun 1y ago
I make sure my management sees how poorly many of the attempts to use LLMs end up going, even for basic stuff it should do well.
261.
by j2kun 1y ago
If the last time you saw a wrong answer was a year ago, then you are definitely regularly getting them and not noticing.
262.
by j2kun 1y ago
Fail them enough and it will sink in I'm sure.
263.
by j2kun 1y ago
> This is where the hallucination began The funny thing is that is also "hallucinates" when it does what you want it to do. <insert always has been meme>
264.
by j2kun 1y ago
He also had some interesting things to say about these IMO results: https://mathstodon.xyz/@tao/114881418225852441
265.
by j2kun 1y ago
That is not what they mean by contributing to scientific discovery.
266.
by j2kun 1y ago
> the entirety of that data needs to be loaded into my program What? No. I'm not saying the entire Google search index is feasible, but you can do a lot. Here are some concrete numbers from what is now considered an "old"
267.
by j2kun 1y ago
I don't think this is possible with FHE alone.
268.
by j2kun 1y ago
Comparisons can be implemented by approximating a < b with 0.5 * (sign(a - b) + 1) And the sign function can be approximated by a polynomial that uses only additions and multiplications and products with constants. Other FHE sch
269.
by j2kun 1y ago
> Can they still train the models over encrypted data? Yes but then the model becomes encrypted. IMO ML training is not a realistic application for FHE, but things like federated training would be the way to do that privately enough.
270.
by j2kun 1y ago
You might enjoy https://jeremykun.com/fhe-in-production

More ›