riku_iki

31 ms·

631.
by riku_iki 2y ago
> You just have to think it stands to be the "next" Google, which feels increasingly plausible. Google's current market capitalization is in the trillions. for this, in addition to "google" part, they also need t
632.
by riku_iki 2y ago
The idea is that dev packages dependencies in lib folder, so when people retrieve sources, they also retrieve all dependencies. But you are right, that this is some extra work for dev, so this is question of trade offs: do you want to deal
633.
by riku_iki 2y ago
> selling signals could this have any legal risks, e.g. your clients will sue you for manipulations and causing losses to them? Is this a valid business at all from SEC standpoint?
634.
by riku_iki 2y ago
You can always use something as simple as Make for your C++ proj with manually dumping dependencies to some libs folder.
635.
by riku_iki 2y ago
this article has several examples: https://blog.polybdenum.com/2021/08/09/when-zero-cost-abstra...
636.
by riku_iki 2y ago
> But I'd say nowadays that learning to use Rust well is much easier than learning to use C++ well . For someone(me) who was making a choice recently, it is not that obvious. I tried to learn through rust examples and ecosystems,
637.
by riku_iki 2y ago
> No it didn't. They tested up to 100 digits with very high accuracy. I don't think you even read the abstract of this, nevermind the actual paper. they have two OOD (out of distribution) accuracies in the paper: OOD: up to 100
638.
by riku_iki 2y ago
In my books what they built (specialized training data + specialized embeddings format) is exactly specialized model. You can disagree of course and say again that I don't understand something, but discussion will be over.
639.
by riku_iki 2y ago
What you are saying contradicts to my reading of publication.
640.
by riku_iki 2y ago
> https://arxiv.org/abs/2405.17399 they built specialized model which is after bunch of trickery still has 99% accuracy(naive model had very low accuracy) on very simple deterministic algo. I also think most of the
641.
by riku_iki 2y ago
It could have in order to power AI and Crypto..
642.
by riku_iki 2y ago
> Because anyone who has said nonsense like "LLMs by their very nature cannot do x" and waited a few years has been wrong. That's why GPT-3 and 4 shocked the research world in the first place. there are some benchmarks whi
643.
by riku_iki 2y ago
> Logic programming? AI until SQL came out. Now it's not AI. logic programming is not directly linked to SQL, and has its own AI term now: https://en.wikipedia.org/wiki/GOFAI .
644.
by riku_iki 2y ago
> You will not be able to get that performance out of Postgres if writes are batched, I get this and higher performance from postgres. If 800k on 64 cores is Scylla's best result, it is not that impressive. But also you probably mea
645.
by riku_iki 2y ago
> Scale certainly wouldn't be it. vanilla postgres can't scale to such size, you need some sharding solution on top, which likely will be much harder to maintain than ScyllaDB..
646.
by riku_iki 2y ago
there are no much reliable benchmarks which would measure what is gap really. I think currently corps compete in who will leak benchmarks to training data the most, hence o1 is world programming medalist, yet makes stupid mistakes.
647.
by riku_iki 2y ago
> Serious question: Why aren't Libgen, Annas-Archive, and others operating solely as an onion service on TOR? I think hosting in country outside of US jurisdiction (Russia?) and being accessed through VPN is more consumer friendly s
648.
by riku_iki 2y ago
> You have to admit, there’s something appealing about being able to read the entire spec for a popular, performant, production language over a lunch break. You don’t need to read and/or master a language’s specification in order to
649.
by riku_iki 2y ago
> C, Basic, Pascal, etc. are all missing inheritance and generics, are they similarly unorthodox and difficult? yes, they are unorthodox for this specific niche: enterprise devs, most languages have OOP and most devs know it. > Java,
650.
by riku_iki 2y ago
> but that most code bases will require you to earn an architecture astronaut certification for that particular code base. sure, if go would be popular in enterprise dev, we would see magnitude more error handling spagetti and genericles
651.
by riku_iki 2y ago
> Java's problem is that it doesn't do this well for anything but heavy business logic applications, and the more clever the programmer, the more they can build up complex structures Java has very simple OOP constructs (classes
652.
by riku_iki 2y ago
> Somehow i feel sad for this AI model. it will be the first AI CEO.
653.
by riku_iki 2y ago
you can pick existing pretrained foundational model from corp (google, MS, Meta) and then finetune it(much cheaper) with your innovative ideas.
654.
by riku_iki 2y ago
not 7 digits. But I observe some 7 digit people. They do well pretending being busy though, but it is not relevant to value they deliver.
655.
by riku_iki 2y ago
stress is your internal state, and is optional. Work intensity is result of you not saying no. I still believe if your employer pays you 7 digits for 12h/d, he will pay you about the same for 8h/d which will bring total -20% deliv
656.
by riku_iki 2y ago
> such comp and work life balance are basically incompatible. > comp like this requires serious results there is diminishing return in results once you shift from good work/life balance schedule. Say you work 8h/d and delive
657.
by riku_iki 2y ago
From article: he also created offline/private set of questions to avoid answer leakage to o1 training set, and o1 scored around 97, which is not 120, but still significant leap in LLM performance.
658.
by riku_iki 2y ago
> Remarkably, constant depth is sufficient. I think article also says log(n) embedding size (width?) is required, where n is size of input.
659.
by riku_iki 2y ago
> Meanwhile, in the real world, I've found that I haven't written a line of code in weeks. Just paragraphs of text that specify what I want and then guidance through and around pitfalls in a simple iterative loop of useful work
660.
by riku_iki 2y ago
and what prompt you gave them to generate program? Did you tell explicitly that they need to fill cornered cells? If yes, it is not what benchmark is about. Benchmark is to ask LLM to figure out what is the pattern. I entered task to Claude

More ›