estearum

30 ms·

841.
by estearum 4mo ago
This train of thought is hilariously rife with selection bias. There is an infinitely long list of technologies that "Luddites and doomers" tried to block and were thankfully successful in doing so. CFCs and ozone. Leaded gasoline
842.
by estearum 4mo ago
I don't think this is a problem with economists . It's a problem with incentives and the systems we use to balance those incentives. You have it totally backwards lol. You can remove the economists or even the word "econom
843.
by estearum 4mo ago
I don't hate AI lol. Maybe you should stop making arguments up in your head and, again, address the argument put in front of you. It's incredibly lazy to group technologies by "were people afraid of this? Yes/No" Yo
844.
by estearum 4mo ago
When it comes to what actually gets encoded into reality, it certainly is the most salient factor
845.
by estearum 4mo ago
Excellent work responding to the argument put in front of you! You are so rational!
846.
by estearum 4mo ago
Obviously there are components of nuclear energy that are not intrinsically safe, which is why people freaked out about it
847.
by estearum 4mo ago
In every sense that economically matters (i.e. in every sense that will actually attract the resources required to realize such a determination), it can be.
848.
by estearum 4mo ago
Not really. Nuclear power (in both the weapon and energy sense) is and always was relatively easy to manage compared to AI. If the deployment of nuclear energy required the distribution of radioactive material to every smartphone in America
849.
by estearum 4mo ago
Sure, here's a direct answer: Even after the mechanical loom was created, figuring out the next most important problem to work on was a job that humans did. Unless you believe there's some hard limit on AI intelligence that will
850.
by estearum 4mo ago
Just because people are angry does not make it false. > Why da f*#$ do they have to continue developing a technology which they think will replace droves of people by machines?? Because arms race. Game dynamics and incentive structures d
851.
by estearum 4mo ago
The problem with this method is it will flip immediately upon new Trump utterances. See David Sacks having an absolute meltdown over Bernie proposing some type of partial nationalization of AI labs, then just a few days later radio silence
852.
by estearum 4mo ago
So?
853.
by estearum 4mo ago
Information asymmetry
854.
by estearum 4mo ago
And their mandate obviously affects life and death decisions at massive scale across the US and entire world, silly.
855.
by estearum 4mo ago
Right... which is (and you're never going to believe this) what Anthropic is asking for
856.
by estearum 4mo ago
This is how games are meant to operate. But the balance space is now so incredibly high-dimensionality that it turns out there is almost always a meta-optimal strategy. There is a Rock that beats both Scissors and Paper at greater frequen
857.
by estearum 4mo ago
Reasonable people freak out when they see that on a freeway There are also plenty of unreasonable people who freak out when they see that on a not-freeway
858.
by estearum 4mo ago
No, I have never bought a subscription and then expected it to get arbitrarily upgraded any time a higher tier was introduced.
859.
by estearum 4mo ago
Does your local store sell aerosolized anthrax?
860.
by estearum 4mo ago
Did Anthropic say "this is going to destroy the world, sign up now"? The details matter quite a lot actually in free speech cases. Should a nuclear energy startup that says "nuclear weapons are very dangerous and should be re
861.
by estearum 4mo ago
That's not what happened. DoD signed a contract with Anthropic that said it can't use Claude 1) to run a fully-autonomous killchain or 2) to surveil Americans. DoD then decided they did not want to abide by the contract they sign
862.
by estearum 4mo ago
Just like when you buy a top of the line camera or car, and then they release a new one, you are entitled to the now-top-of-the-line camera or car. What the heck come on.
863.
by estearum 4mo ago
I'm not a fan of age verification(++) discussed here, but it's also extremely obvious that social media is a more significant danger to our society (children and non-) than gas station drugs.
864.
by estearum 4mo ago
100% I vaguely remember some factoid about how Magnus could tell you from memory the 20 exact opening and 20 exact closing moves of every championship game since the dawn of time or some such whatever, and I was like "oh wow great for
865.
by estearum 4mo ago
Yes, I am aware they exist. But the addition of an industry that systematically finds and disseminates this obviously accelerates the process of discovery and dissemination. This is like saying "scientific progress exists even without
866.
by estearum 4mo ago
If the intent was for everyone to use the same build, then game designers would not invest in the concept of builds. The convergence is clearly not intentional (it is a flaw in the game design) but also basically impossible to avoid in the
867.
by estearum 4mo ago
Yes, it is a game design failure (obviously) insofar as it seems not possible to make a perfectly balanced game. But games weren't previously perfectly balanced either, presumably. But they didn't have this same "herding&qu
868.
by estearum 4mo ago
The FDA regulates which drugs make it to market which is itself an extremely powerful force (excessively so) on conversations between you and your doctor. In the case of vaccines, FDA's decisions can quite obviously make a difference
869.
by estearum 4mo ago
One way they should balance in a functioning society is that while tenure would protect you from negative repercussions within the walls of an academic institution, a Congress with any semblance of seriousness and care toward the American p
870.
by estearum 4mo ago
No, you are confounding what I'm talking about (fast discovery and dissemination of meta-optimal strategies) with toxicity. You could have 0 toxicity at all and still have a game totally ruined by the behavior I'm pointing out.

More ›