lumost

29 ms·

571.
by lumost 2y ago
I believe Warren Buffet was opposed to robo-trading strategies for this exact purpose. If the bulk of the money is going to fund anything with a market cap greater than $X, then it is useful for VCs to pump a stock up to $(X + Y) market cap
572.
by lumost 2y ago
anecdotally, I'm flipping back and forth between o1 and GPT-4. o1 is mildly better at editing larger code segments. I worked with it to edit a large ~2k line python file in an unusual domain. But o1 is also incredibly verbose. It'
573.
by lumost 2y ago
That is also much lower than Uber at its peak.
574.
by lumost 2y ago
An open small model means I can experiment with it. I can put it on an edge device and scale to billions of users, I can use it with private resources that I can't send externally. When it's behind an API its just a standard margi
575.
by lumost 2y ago
A high-quality early stage team that self-selects out of follow-up rounds may be a decent outcome for some VCs. This means early liquidity in all of the "positive" events. If the founders were high quality, spinning an acquihire o
576.
by lumost 2y ago
It’s useful to consider the rise of computer graphics and cgi. When you first see CGI, you might think that the software is useful for general simulations of physical systems. The reality is that it only provides a thin facsimile. Real si
577.
by lumost 2y ago
6.5 Billion won't move NVidia's needle at this point. You need exponentially growing revenue from AI firms to drive increased spend on compute.
578.
by lumost 2y ago
Possible the people writing those checks want more experienced people in those positions, or their own people. Or the execs figure a ~150x growth ride in 6 years is about the best that they can hope for and are ready to bounce. If I were in
579.
by lumost 2y ago
The risk is that the combined market caps of AI firms exceed 1 Trillion dollars now. This market cap is likely supported by several Billion dollars in Revenue, but margins are debatable outside of NVidia due to capex/opex of inference,
580.
by lumost 2y ago
I suspect that the motivating factor is lack of deployable human capital. In a large centralized society, there may always people to deploy at the problem. The more people you deploy at the problem, the more time the people at the top spend
581.
by lumost 2y ago
To my knowledge, this has only ever been done successfully during ww2. The mechanism used was semi-forced lending via war bonds. The war bonds were predominantly purchased by wealthy individuals and the rate of return was lower than all alt
582.
by lumost 2y ago
Out of curiousity, what do you dislike about spreadsheets/google docs? This has been my primary means of tracking progress historically. I've tended to find that all other mechanisms just add unecessary overhead.
583.
by lumost 2y ago
All medical advice is inevitably biased by who doctor's see - not who they don't see. Pediatricians encountered peanut allergies, so they took action - cardiac patients appear with high blood pressure, so action is taken. Advice a
584.
by lumost 2y ago
Curious what other items end up there. E.g. particulate dust, dirt, minerals etc. It’s a more interesting result if plastics are unique in ending up here.
585.
by lumost 2y ago
This is the reason that consultants often get paid ridiculous hourly rates, and in-demand consultants may overbook. So far, my experience has been those that are likely over-employed - tend to be poor workers. I might be "ok" with
586.
by lumost 2y ago
I spent some time understanding early stage VC back in 2022/2023. My anecdotal experience is that there is a lot of hesitance by most funding sources to go after bigger problems. You tend to get a lot of advice along the way to focus o
587.
by lumost 2y ago
Likely an attempt to break out of the mold. It's possible to adjust YC style incubators to work on harder problems that require more capitalization - but the milestones look different. You wouldn't encourage a nuclear reactor star
588.
by lumost 2y ago
Sure, but usually these things can be accelerated if you are blocked and can't do other work. It's usually expected that you are doing other work while waiting, be it ops, reviews, invention/simplification, design, coding for
589.
by lumost 2y ago
The concern with over-employment is that many "healthy" organizations rely on trust. Someone says it takes ~4 weeks to do something, I don't want to have someone else "re-scope" the effort to verify that it really t
590.
by lumost 2y ago
on average, everyone will follow the path of maximum reward/lowest risk. Management has a job to do, but is aware that they may lack any choice or path to success. They still have a job to do in e.g. allocating engineers work, managing
591.
by lumost 2y ago
LLMs perform well on small tasks that are well defined. This definition matches almost every task that a student will work on in school leading to an overestimation of LLM capabiity. LLMs cannot decide what to work on, or manage large bodie
592.
by lumost 2y ago
There are multiple ways to run a business like this. 1. Go deep on the tech, there are funders who will want equity stakes in risky startups because they operate in adjacent markets. It's often cheaper to invest 1MM on a startup than i
593.
by lumost 2y ago
Or companies used that as a cover for gouging cuz everyone else was doing it. There are brands like Arizona which did not change prices. Odds are that the truth is somewhere in between.
594.
by lumost 2y ago
Making the entire thing efficient enough to actually be delivered to a target is also another matter. This requires precise calculation of the geometries and very precise grades of plutonium, barrel pipe, and explosives. How do you even kee
595.
by lumost 2y ago
A 26% productivity increase is... pretty big? There is obviously some noise in the data. But given copilots are pretty new pieces of tech, it does not seem impossible that there could be at least a 2x productivity growth available within th
596.
by lumost 2y ago
Licensing means recurring payment, why not buy elsevier outright?
597.
by lumost 2y ago
If the model weights are distributed, or its a free forever product... Fair use seems like a solid claim for LLMs in my semi-informed opinion. It seems like it will be a tall leap to create a trillion dollar corporation as fair use. It woul
598.
by lumost 2y ago
SpaceX's finances are also largely unknown, they could be burning more money than NASA on experimental flights.
599.
by lumost 2y ago
Anecdotally, I've never seen a project be a career ender. Most individuals have many projects under their belt, it's rare that an individual has bet the farm on an effort in such a way that others would not employ them. When a pro
600.
by lumost 2y ago
I agree with you on the overall point, but this statement. > Not possible. In all honesty, Cosmology rests on the principal that physics is the same in all directions, over all translations, and over time translation. While this is a goo

More ›