j2kun

10 ms·

211.
by j2kun 1y ago
Came here to say this, but with caveats. The particular domain has extra properties that allow their "stupider" algorithm to work better in their case. But a general graph drawing system has to deal with the inherent generality of
212.
by j2kun 1y ago
Tell that to the guy whose wife cheated on him with an LLM
213.
by j2kun 1y ago
Perhaps to muddy future corpora
214.
by j2kun 1y ago
balajis is personally invested in propping up the ecosystem.
215.
by j2kun 1y ago
They wouldn't be fine with it if they understood it. Instead they are not paying attention and only consuming propaganda media (Fox News) that refuses to report on Trump negatively in any way.
216.
by j2kun 1y ago
Why did Google decide to choose Mercurial? Based on what I read the main reason was that the mercurial dev team was willing to prioritize features needed for Google to add custom extensions to support its monorepo, and the git dev team wasn
217.
by j2kun 1y ago
I started using git as my daily driver at Google in 2018. I was forcibly kicked off it for mercurial in 2020. The deprecation plan for mercurial is 2026 IIUC. Maybe if you adopted mercurial early it's different.
218.
by j2kun 1y ago
> I also don’t mean to imply that everyone at Google is using jj, but the contingent feels significant to me, given how hard it is to introduce a new VCS inside a company of that size. I don't mean to imply that Google is fickle, bu
219.
by j2kun 1y ago
> everyone is deathly afraid of AGI I think this is a vast overstatement. A small group of influential people are deathly afraid of AGI, or at least using that as a pretext to raise funding. But I agree that there are so many more things
220.
by j2kun 1y ago
What overwhelming evidence do fossil records provide about human cognition?
221.
by j2kun 1y ago
Even if you don't believe in God, scientific theories of how human cognition came about (and how it works and changes over time) are all largely speculation and good storytelling.
222.
by j2kun 1y ago
> Engineers design and build things on top of the discoveries of scientists I agree with a lot in this post, but I think it's also worth mentioning how this is a two-way street. Practical considerations often drive theory research a
223.
by j2kun 1y ago
Polytechnic schools seem to do this well, research universities not so much.
224.
by j2kun 1y ago
And yet, isn't antirez fully on board with LLMs now?
225.
by j2kun 1y ago
> Today’s real chain: React → Electron → Chromium → Docker → Kubernetes → VM → managed DB → API gateways. Each layer adds “only 20–30%.” Compound a handful and you’re at 2–6× overhead for the same behavior. > > That's how a C
226.
by j2kun 1y ago
> The empirical argument > We can ask a question: how long (in nanoseconds) does it take to access a type of memory of which an average laptop has N bytes? Here's GPT's answer: "Here's what GPT says" is not an
227.
by j2kun 1y ago
Sheldon Axler's book is the common (now decades old) example of a book doing linear maps first.
228.
by j2kun 1y ago
+1 this is a great reference text
229.
by j2kun 1y ago
I wrote one: https://pimbook.org
230.
by j2kun 1y ago
Politics is any question of the form "what should we do?" If you don't want to be involved in answering questions like that, then by all means avoid politics.
231.
by j2kun 1y ago
This seems like the opposite of "privacy by design" > Privacy by design: Your reading habits belong to you. We don’t track, profile, or monetize your attention. You remain the customer and not the product. But the person runnin
232.
by j2kun 1y ago
My guess (paywall): they wanted to drive up demand for their existing investments in AI. In other words, the bubble didn't eat Y combinator, Y combinator did it to themselves on purpose.
233.
by j2kun 1y ago
The Big Bang Theory is not "proved"... From the Nobel Prize press release linked in the article: > it is thus tempting to assume that the universe was created by a cosmic explosion, or ‘big bang’, although other explanations ar
234.
by j2kun 1y ago
It is not that bad these days, closer to 10,000x. Accelerators are being developed that claim to get down to 10x, though i think they will be more like 100-1000x, which would still be a huge improvement considering how people use LLMs today
235.
by j2kun 1y ago
LLMs are at the current forefront of FHE research. There are a few papers doing some tweaked versions of BERT in <1 minute per token. Which is only ~4 orders of magnitude slower than cleartext. https://arxiv.org/html&#x2F
236.
by j2kun 1y ago
Exactly. If you aren't willing to pay the costs, then deal with the terms.
237.
by j2kun 1y ago
The US government can train their own damn LLM if they want an unrestricted one so bad.
238.
by j2kun 1y ago
Except, you know, the dollar bills the government itself prints.
239.
by j2kun 1y ago
Every company that does web scraping has one. My point is that it's not about RICHNESS (what can be done) because everything mentioned by the OP is already done in C++.
240.
by j2kun 1y ago
You made my point for me: DuckDB is already in C++ (so C++ is rich enough), but hard to glue with other pieces, and you don't want to put in that effort, so you use Python.

More ›