noosphr

29 ms·

721.
by noosphr 1y ago
If it were a good metric there wouldn't be a few phone books worth of regulations on what you can do before and during running 100 meters. From banning rocket shoes, to steroids, to robot legs the 100 meter run is a perfect example of
722.
by noosphr 1y ago
All my pcs have a locked down guest/guest account which gets nuked on logout.
723.
by noosphr 1y ago
OP probably had to update Windows today. It's a weird world where KDE out of the box is more usable than Windows. I gave the wrong laptop to my mother a few months back and she only told me when she finished banking that the windows me
724.
by noosphr 1y ago
The same way that we dealt with Excel programming. Ignore it until it blows up, then spend hundreds of thousands trying to fix it before the company goes bankrupt.
725.
by noosphr 1y ago
This is akin to complaining about the overhead of radix conversion between decimal and binary in early computers. Unless you figure out how to talk in latent space directly to an llm the preamble is the cost you pay for getting the answer b
726.
by noosphr 1y ago
You can quite easily ask it to summarize the result in a sentence or paragraph. LLMs have no other way to compute than write text and the more text they write the more compute they do. You only care about the final output.
727.
by noosphr 1y ago
Steam power started with atmospheric engines which didn't need steel.
728.
by noosphr 1y ago
I read these articles about how much of a suckers bet AI is and start feeling existential dread about the coming market crash. Then I remember I just did a weeks worth of work in a few hours and feel a lot better about my prospects at least
729.
by noosphr 1y ago
I've found that xml is surprisingly good for llms when it comes to table extraction in production. I only found out when I send the raw xml storage format to benchmark again various flavours of everything else. XML turns out to the bes
730.
by noosphr 1y ago
Mailing lists. They've been around since the 80s. They solve all these problems. They are amazing. Use them.
731.
by noosphr 1y ago
Both parties like it better because it turns the electricity market into another casino that lets you take billions from the parts of the economy that do things. I worked as a quant in the electric market. There wasn't a single dataset
732.
by noosphr 1y ago
Malthusianism for computers. Moore's law is dead. The only way to increase compute is to increase the power we feed to computers. AI is just the shiniest example. Everything else will follow suit until electricity costs increase enough
733.
by noosphr 1y ago
The power bill going up is because the US, and the West in general, bet on renewables and a low energy future. Neither of those things turn out to be a good fit for the new economy. The only thing left for people who derided nuclear for the
734.
by noosphr 1y ago
You can email them and they are usually quite happy to talk. Of course not when they get a media storm like these people. But I regularly correspond with experts in adjacent fields who have interesting papers put out.
735.
by noosphr 1y ago
Does anyone have the pre-print? I'm not affiliated with a university any more and the usual suspects don't upload papers overnight any more.
736.
by noosphr 1y ago
Drop me a line on my profiles email. I'm playing around with using hyperlinks in pdfs to get around how much the www sucks for posting serious research with serious working code. Caveat emptor: I'm first working on getting the bas
737.
by noosphr 1y ago
I have no idea why you're bringing up the Turing test, then saying it doesn't matter when you used it wrong. What I'm talking about is that I use Claude daily to find stupid bugs in my code. I can now spend my time on things
738.
by noosphr 1y ago
Sounds like you need more Emacs in your life. This post was written in Emacs.
739.
by noosphr 1y ago
> His books, many of which are annotated with margin comments, I'm not saying that he did, but this along with being the right age to have read How to Read a Book by Mortimer J. Adler strongly suggest that he used that book to grasp
740.
by noosphr 1y ago
It did not.
741.
by noosphr 1y ago
I can talk to a computer and it can talk back. The rest is commentary.
742.
by noosphr 1y ago
No one expected natural language to be solved in 2020. At the time the smart money was on 20 to 80 years. Many people are still acting like it's not. Things happen not at all and then all at once. We have the cheap humanoid platforms c
743.
by noosphr 1y ago
Natural language wasn't solved by brute force until we started using trillion parameter models and using the whole internet, every book and every article ever published as training data. I don't know of anyone spending tens of bil
744.
by noosphr 1y ago
I have an n9 that I turn on once in a blue moon. Even today it looks better than any of the major mobile OSs, is more responsive and just _feels_ better to hold and use. The road not taken.
745.
by noosphr 1y ago
Mailing lists are still the superior format. The barrier to entry is a feature and not a bug.
746.
by noosphr 1y ago
>The article sort of goes sideways with this idea but pointing out that AI coding robs you a deep understanding of the code it produces is a valid and important criticism of AI coding. You can describe what the code should do with natura
747.
by noosphr 1y ago
People act like big tech didn't have a mountain of cash they didn't know what to do with. Each of the big players has around 100 billion that just sitting there doing nothing.
748.
by noosphr 1y ago
Yes, that's been the downside of these forever. If you use quantized differentiation you can get away with using integers for gradient updates. Explaining how takes a paper and in the end it doesn't even work very well. At univers
749.
by noosphr 1y ago
Ed was also line oriented. Using regex to edit lines instead of typing them out was a step up, but not much of one. Also my father definitely had C punch cards in the 80s.
750.
by noosphr 1y ago
Just ask an Ai to write how it made you more productive in daily work. It's really good at that. You can pad it out to 1m words by asking it to expand on each section of with subsections.

More ›