Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
lumost
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
35 ms
·
661.
▲
by
lumost
3y ago
Unless you are NVidia.
662.
▲
by
lumost
3y ago
Has anyone tested how close you need to be to the weights for copyright purposes?
663.
▲
by
lumost
3y ago
This happened to WizardLM2 yesterday as well https://wizardlm.github.io/WizardLM2/
664.
▲
by
lumost
3y ago
I spent many years feeling guilty about this. Particularly during covid when, on paper, there wasn't a lot else going on. Anecdotally, the thing to remember is that you brain is an organ - Even professional trainers can't do their
665.
▲
by
lumost
3y ago
We're fast approaching the point where many events will "happen". Geo-engineering is easy and economical. Small Governments/Billionaires/Corporations will pursue it barring a legal framework to ban it. Given the lac
666.
▲
by
lumost
3y ago
biggest win for me as a senior engineer is the ability to quickly prototype something in a new framework/language. In these cases, I generally know what I want and what "good" looks like - but I'm missing context on the
667.
▲
by
lumost
3y ago
I believe this is overly skeptical. While the tech plainly isn't there yet it's easy to see where things are going. Unlike the voice assistants of old, the newer ai tools can be blended into radically different use cases with
668.
▲
by
lumost
3y ago
Best decision I ever made was to skip the duvets and buy a polar fleece blanket for $30. Not to hot, not to cold, machine washable and $30.
669.
▲
by
lumost
3y ago
My understanding of the games industry is that a small fraction of game programmers are working on core game engines and low level graphics kernels. Is that inaccurate?
670.
▲
by
lumost
3y ago
How hard do you think it is to find engineers who are - Great at legacy C++ code. - Great at new C++ code. - Great at embedded/high performance/distributed code. - Are experts in Linear Algebra and Calculus - Are competent at Mach
671.
▲
by
lumost
3y ago
Also points towards frustratingly limited progress in cancer treatment efficacy.
672.
▲
by
lumost
3y ago
The mass processing of food could be the additional mechanism allowing mass distribution of prions, no? Through processed foods, an apex predator like humanity comes into contact with proteins from several orders magnitude more animals th
673.
▲
by
lumost
3y ago
Doesn't this depend on the STP of the alien environment? It seems plausible that planets that are a few hundred degrees warmer might benefit from silicon based life, no?
674.
▲
by
lumost
3y ago
I've heard this pitch from a few AI labs. I suspect that they will fail, customers just want a model that works in the shortest amount of time and effort. The vast majority of companies do not have useful fine tuning data or skills. Co
675.
▲
by
lumost
3y ago
https://wtfhappenedin1971.com/ Annual wages are about 15% higher today compared to 1971 in real terms. Real terms is a bit suspect in this case as wage earners are more likely to pay nominal rent then owner equivalent rent.
676.
▲
by
lumost
3y ago
This occurs far more often than you expect. When I first started in Tech 15 years ago in a major urban environment - I was extremely surprised to see young professionals living in housing that I viewed as utterly unaffordable. For a time I
677.
▲
by
lumost
3y ago
This also belies the stark age differences between owners and non-owners. Ignoring that it is becoming dramatically more expensive to move from one category to the other does not seem correct. Particularly if the majority of individuals wou
678.
▲
by
lumost
3y ago
Reality is that we likely incentivize that behavior amongst scientists. If tenure comittee just cares about paper count, and the average 2-4 citation paper is never replicated. Then there is a viable strategy for publishing crap work faster
679.
▲
by
lumost
3y ago
Newton could not have made testable predictions about quantum mechanics. There is a good argument to be made that we’ve explored terrestrially accessible energy scales for the time being. It’s hard to make theoretical progress around such a
680.
▲
by
lumost
3y ago
The interesting emphasis to me is commercial customers. They are acknowledging that the competition to be the nth gpt vendor is too stiff for them, and they aren’t successful as an independent venture funded research lab. I can think of 4-1
681.
▲
by
lumost
3y ago
Yeah debugging would be a pain, but in the context of inference/training unnecessary. There is some set of ops which requires high precision, if I L2 normalize a tensor - I really need it to be normalized. But matmul/addition? May
682.
▲
by
lumost
3y ago
It did make me curious however, if we dropped the requirement that operations return correct values in favor of probably correct values - would we see any material computing gains in hardware? Large neural models are intrinsically error cor
683.
▲
by
lumost
3y ago
One tricky area of feedback culture is the question of "is the person open to feedback right now?". In particular, if I'm working more than 50 hours in a week - I stop being open to feedback unless that feedback is "this
684.
▲
by
lumost
3y ago
How many people processes do you currently interact with that "barely work". - Call Customer Service, go through a crazy flow chart and be redirected to 10 different people. - Go shopping, sift through thousands of reviews/im
685.
▲
by
lumost
3y ago
permanently grounding 737 MAX would be tantamount to killing Boeing's commercial aviation products (and possibly the company itself). Unfortunately, this would have the side effect of destroying America's only viable commercial av
686.
▲
by
lumost
3y ago
I saw an ad recently "You'll never code as well as ChatGPT." I suspect we'll see some changes to both how and why people code, and it's entirely likely that application types which are difficult for AI to code will
687.
▲
by
lumost
3y ago
Out of curiousity - how does a VC fund hold onto "powder"? Do they have terms to invest in a liquid fund, or some other arrangement? Seems like they'd face tough returns if they held powder for long.
688.
▲
by
lumost
3y ago
Anecdotally, not much changed for East coast startups. What did change is that West coast firms came to the east coast. Whereas 10 years ago it would be noteworthy to work for an SV company - it is now typical. I think the local market wa
689.
▲
by
lumost
3y ago
> Ahh, yes the marxist slight of hand to turn compounding on foregone consumption into something evil rather than something that should be applauded and appreciated. What's the foregone consumption on 1 Billion+ dollars of inheritan
690.
▲
by
lumost
3y ago
This may explain the substantial performance increase in proprietary models over the last 6 months. It also may explain why open-air and others had to drop open models. Distributing copyrighted material via model weights would be problemati
More ›