lossolo

16 ms·

241.
by lossolo 1y ago
It will not crash, they will just print more money, which will raise inflation and devalue debt. So elites and speculators will make money by pumping everything up with debt, and then they will tax (through inflation) the people who need pa
242.
by lossolo 1y ago
https://www.figure.ai/company "Building Figure won’t be an easy win; it will require decades of commitment and ingenuity." "Our focus is on what we can achieve 5, 10, 20+ years from now, not the near-term win
243.
by lossolo 1y ago
Calling that a "footgun" feels really overstated. Yes, reserve is the safer default because it gives you slack and preserves amortized growth even if your usage pattern is messy. But "needs global analysis" isn't un
244.
by lossolo 1y ago
Sure, but this isn't a job for reserve_exact. Vec::reserve is like a framing hammer made for speed and momentum. You might dent a little extra wood (overallocate), but you drive lots of nails fast and keep amortized O(1) growth. reserv
245.
by lossolo 1y ago
Quadratic pushfive is just a cautionary tale about misusing reserve_exact. Basically use reserve_exact when you know the final size (reserve once), or you're doing a one off tight sizing where memory footprint matters. Don't pre r
246.
by lossolo 1y ago
> Vec::reserve_exact is a more narrow idea - we can hint about the ultimate capacity needed, if we're wrong and later need more capacity this has a significant performance cost because we thew away the amortized growth promise to ge
247.
by lossolo 1y ago
I agree that computational irreducibility and chaos impose hard limits on prediction. Even if an intelligence understood every law of physics, it might still be unable to simulate reality faster than reality itself, since the physical world
248.
by lossolo 1y ago
My point wasn't so much about how fast humans achieved these things, but about what's possible at all given a certain cognitive architecture. Chimpanzees could live for another million years and still wouldn't build a factory
249.
by lossolo 1y ago
Generally, I agree, but it also depends on perspective. Intelligence exists on many levels and manifests differently across species. From a monkey's standpoint, if they were capable of such reflection they might perceive themselves as
250.
by lossolo 1y ago
Interesting, do you have a link to the study? Our experience is different, at least when reviewing LLM generated code, we find quite a few errors, especially beyond 200 LOC. It also depends on what you're reviewing, 200 LOC != 200 LOC.
251.
by lossolo 1y ago
And people forget what was actually licensed and what value does it have without ASML. ASML bought SVG lithography two decades ago to expand in dry/immersion optical litho and US market access. EUV matured later. SVG had momentum in DU
252.
by lossolo 1y ago
In our experience, it depends on the task and the language. In the case of trivial or boilerplate code, even if someone pushes 3k-4k lines of code in one day, it's manageable because you can just go through it. However, 3k lines of int
253.
by lossolo 1y ago
> China has a hyper protected domestic market When I was in China, I saw Teslas everywhere and iPhones too. It seems there are products that can still compete in China against strong domestic brands. The country hasn't really been h
254.
by lossolo 1y ago
What's the security story in Ladybird? Do they use sandboxes etc? I'm a bit concerned that hundreds of thousands of lines of C++ browser code written in just three years could be a minefield, but I hope I'm wrong.
255.
by lossolo 1y ago
Ye, that's basically an MCP server, that can be used by ChatGPT.
256.
by lossolo 1y ago
> Porsche 919 Hybrid EVO did it in 5:19 If anyone hasn't seen this, I highly recommend it, even if you're not a car fan. https://www.youtube.com/watch?v=PQmSUHhP3ug
257.
by lossolo 1y ago
In an interview with Politico in 2023, former Israeli Prime Minister Ehud Olmert said that "In the last 15 years, Israel did everything to downgrade the Palestinian Authority and to boost Hamas." He continued saying "Gaza was
258.
by lossolo 1y ago
This, for Israel it was a perfect pretext. Netanyahu financed Hamas through Qatar for years because he knew that Hamas is a destructive force for Palestinians and their struggle for nationhood. It was a cold, cynical decision. He exploited
259.
by lossolo 1y ago
The funniest thing about all this is that the biggest difference between LLMs from Anthropic, Google, OpenAI, Alibaba is not model architecture or training objectives, which are broadly similar but it's the dataset. What people don&#x2
260.
by lossolo 1y ago
Poland specifically bought Pegasus from Israelis. There is a parliamentary commission investigating this, everything is public so you can google it. Poland does not develop its own exploits or have domestic spyware firms that target journa
261.
by lossolo 1y ago
Most hybrid stacks (BM25 + dense via HNSW/IVF) still rely on embeddings as a first class signal. So in practice the vector side carries recall on paraphrase/synonymy/OOO vocab, while BM25 stabilizes precision on exact term an
262.
by lossolo 1y ago
Sure, but vector search is the dominant form of RAG, the rest are niche. Saying "RAG doesn’t have to use vectors" is like saying "LLMs don't have to use transformers". Technically true, but irrelevant when 99% of wh
263.
by lossolo 1y ago
When I was in Shenzhen, all deliveries to hotel rooms in our skyscraper were done by autonomous robots. The delivery person would put the food or other items into the robot in the lobby, select the hotel room, and the robot did the rest, fr
264.
by lossolo 1y ago
RAG is a technique, so instead of string matching (like grep), it uses embeddings + vector search to retrieve semantically similar text (car ≈ automobile), then feeds that into the LLM. Tool use is broader, RAG is one pattern within that, b
265.
by lossolo 1y ago
In RAG, you operate on embeddings and perform vector search, so if you search for fat lady, it might also retrieve text like huge queen, because they're semantically similar. Grep on the other hand, only matches exact strings, so it wo
266.
by lossolo 1y ago
https://ec.europa.eu/eurostat/statistics-explained/index.php... https://ec.europa.eu/eurostat/statistics-explained/index.php...
267.
YouTube settles with Trump for $24.5 millon(politico.com)
11 points by lossolo 1y ago|3 comments
268.
by lossolo 1y ago
> They have hundred millions of paying subscribers They have hundreds of millions of users in total (free tier included), with around 10-15 million paying users.
269.
by lossolo 1y ago
The same reason YouTube reviewers and influencers get access to hardware or games before release. In this case, the person is a passionate blogger.
270.
by lossolo 1y ago
> If you, like the whole world, consume Redis through a network connection I think "you are definitely missing some key insights onto the topic". The whole world is a lot bigger than your anecdotes.

More ›