staunton

28 ms·

421.
by staunton 3y ago
I'm absolutely positive a few million dollars could get you a system that can "read aloud" pdf math papers in no time. I guess people will wait for it to become cheaper though.
422.
by staunton 3y ago
Just to drop the name, we're talking about quantum key distribution (QKD). > Now, what bits do you actually send? You just send random bits and get a key. Then use the key however you want (AES, one-time-pad if you're crazy...)
423.
by staunton 3y ago
The hype might end by that time. Maybe it will be "self-replicating self-organizing microrobots" by then...
424.
by staunton 3y ago
Can you recommend something high quality?
425.
by staunton 3y ago
> the usual pragmatism applies That sounds like by "follow" you don't mean "study a lot"... What do you mean then?
426.
by staunton 3y ago
> Large Language Models are the efforts of mapmakers to encapsulate reality in a way that is ultimately futile. What's futile? Nobody is trying to "encapsulate all of reality". Trying to do that would be futile but succeed
427.
by staunton 3y ago
Arguably, biochem also has "solid foundations" in physics. Would you recommend someone in your field to "follow the physics" in their journey?
428.
by staunton 3y ago
This is true for people who have worked on big projects or had to maintain working software over time. However, I dare you to look at the kinds of code that "Jupyter notebook only needs to run once"-scientists write... Many of the
429.
by staunton 3y ago
What's even more tricky is that almost all (successful) projects start out small and it takes immense discipline and foresight to catch up on "maturity" at the right stage in the project. Chances are, you'll get bogg
430.
by staunton 3y ago
Just buy some datasets
431.
by staunton 3y ago
> I periodically check out YouTube while not signed in I wonder if that actually works... It still tracks location, obviously Cookies (which you probably delete for this) and possibly a lot of other things. Maybe YouTube even maps your I
432.
by staunton 3y ago
Being same as you in terms of Twitter use, I see people talk about it a lot more now (mostly discussing the sudden quirky changes).
433.
by staunton 3y ago
In my experience that's false.
434.
by staunton 3y ago
> almost never Having agreed with everything up to this point, that's a rather bold claim. I would understand it as "supply and demand are not the primary factor in determining almost all prices". Is this really your cla
435.
by staunton 3y ago
You can just define your own malloc and free in any C program. Am I missing what you're talking about?
436.
by staunton 3y ago
Technically, no compiler ever implemented "all" of c++11... You'd have to implement garbage collection, for example :D
437.
by staunton 3y ago
I don't know. How do they?
438.
by staunton 3y ago
> We all know all this "AI safety" is for show, right? No. A lot of people think it really matters A lot of other people pretend to care about it because it also enables stifling the competition and attempting regulatory captur
439.
by staunton 3y ago
Indeed, making poor people in 3rd world countries rate the worst sludge of the internet for 8+h a day might backfire on your marketing... OpenAI could risk it, Google maybe doesn't want to...
440.
by staunton 3y ago
> It exists in a physical (if distributed) location and requires inputs and outputs in electric cables and fibre cables "Nuke all of civilization right now!"
441.
by staunton 3y ago
Effective and manifested power cannot be reined in. A power that is reined in is only as powerful as its reined in form. Most likely, there wasn't any power to start with and it's just a stupid conspiracy theory... You can imagine
442.
by staunton 3y ago
> I think cybersecurity is hard because we are uniquely bad at evaluating cybersecurity risk It's not that hard to evaluate. The risk is high! Pretty much anything can be hacked if you try hard enough... What's hard is everythi
443.
by staunton 3y ago
An organization's policies are still implemented and maintained by humans. No matter how powerful, the power is transient. You have churn, corruption, incomplete knowledge transfer, etc. AI systems in effective leadership could be able
444.
by staunton 3y ago
I'm personally least worried about short-term unemployment resulting from AI progress. Such structural unemployment and poverty resulting from it happens when a region loses the industry that is close to the single employer there a
445.
by staunton 3y ago
It's obvious to me that humans will eventually go extinct. So what? That doesn't mean we should stop caring about humanity. People know they themselves are going to die and that doesn't stop them from caring about things or m
446.
by staunton 3y ago
> Any AGI that is a threat to humanity has to be suicidal to act on it because of how the world works A "threat to humanity" need not mean "killing all humans next year". It could just mean some entity gaining permane
447.
by staunton 3y ago
Most people believe they themselves cannot be influenced by ads "at all". If they were right, nobody would bother making ads.
448.
by staunton 3y ago
It's plausible that in many settings this is actually most efficient from the point of view of the company. Do training on the things that can cause trouble longer term (inconsistent tooling taking over, security issues, company-specif
449.
by staunton 3y ago
> yet they refuse That doesn't seem like a contradiction at all. As you say, they want to hire trained workers rather than providing training themselves. Also, complaining is free while hiring good people is not...
450.
by staunton 3y ago
> I don't think anyone is advocating we dump more sulphur dioxide into the air, People most definitely are advocating that...

More ›