gfodor

39 ms·

481.
by gfodor 4y ago
You have it backwards, you have to calculate the expected value of the mission and see if it’s less than 1.5M. For example, if odds of success are one in 10k for a discovery worth $1T then it works out.
482.
by gfodor 4y ago
$1.5M ain’t much depending on your priors and how much you value the discovery of extraterrestrial tech.
483.
by gfodor 4y ago
I think the question of "deployment" is an interesting one when it comes to smart contracts. When you deploy a contract, you're publishing the code to the network to run, not deploying it to a server you operate. I think we a
484.
by gfodor 4y ago
That's not how universal rights work. You can say privacy is not a universal right. But if it is, governments have no say on if it ever "ends."
485.
by gfodor 4y ago
It has been wild watching this crowd slowly morph over the past decade or so, largely, into exactly what feels like this.
486.
by gfodor 4y ago
> think a simpler explanation is that they have good reasons to believe that this guy was doing more than just the stuff you're trying to defend This isn't a simpler explanation than the even simpler one that goverments have d
487.
by gfodor 4y ago
I mean, we can agree to disagree if you think there would be quiet acceptance of a law where it is criminal to search for cryptocurrency information in google.
488.
by gfodor 4y ago
You have to keep running that thought experiment. If a computer search or errand text resulted in criminal charges, then there would be an opposing counterforce to that. I don't know how it ultimately would sift out, but I don't t
489.
by gfodor 4y ago
It's a DAO. The idea of most DAOs is they are set up to grow into being more decentralized and driven by smart contracts as the number of stakeholders goes up. I am not familiar with the details of how MakerDAO is structured but the fa
490.
by gfodor 4y ago
I don't think this is the same thing. There are technologies now within crypto that are outside of the ability of the US government to nuke from orbit, like the Bitcoin blockchain. The frontiers of DeFi and DAOs however still have no
491.
by gfodor 4y ago
Wow, no offense, but this is an insane level of mental hoops to jump through to insult people who are putting forward the MOND hypothesis. It’s almost invisible, but I think it’s there.
492.
by gfodor 4y ago
That’s just one frame on the question at hand, that tilts the poll in one direction. The principle at hand here is if online transactions ought to have a floor on visibility by law enforcement higher than using cash. So there are other vers
493.
by gfodor 4y ago
Haha I read “sanctioned” as in “endorses” and was very, very confused.
494.
by gfodor 4y ago
I thought this too but these holes don't look very engineered. They look random or organic. That said, this would be an interesting way to create an 'existence proof' of technological capability and presence if there was some
495.
by gfodor 4y ago
Well maybe then it only proves people don’t read the articles they comment on :)
496.
by gfodor 4y ago
I've been downvoted but what you state is exactly right: they have to decide exactly how confident they are in their own process holding up in court (which isn't exactly the same thing as it being accurate) - and to whatever degre
497.
by gfodor 4y ago
The article’s point is basically proven by how many people here don’t even understand he’s talking about this, and not killing the program with Ctrl-C.
498.
by gfodor 4y ago
The point to me seems to be to set up things where this won’t go to court: if the primary claim is that Twitter misstated mDAU, the downside risk to Twitter is far greater than to Musk (I think) to have that fact finding occur. If it turns
499.
by gfodor 4y ago
Yes I think you should spend some time cleaning up your PoC into a real, usable project. Write a nice readme, get it so anyone can run it quickly and easily, etc. In my most recent github project I released [1] I also used DALL-E to make a
500.
by gfodor 4y ago
DALL-E either doesn't generate images you want to see, or if it does, it does a bad job, because it generates many images you don't want to see. In other words, the claim you've set up is basically unfalsifiable, given that t
501.
by gfodor 4y ago
Use twitter long enough and it's hard to not be convinced that p-zombies exist - it may be that such zombies are ephemeral though, and a transitory state of some individuals.
502.
by gfodor 4y ago
Not exactly. If you're using the word "simply" you probably don't quite understand what exactly DALL-E is doing. It's not just mapping correlations between words to pixels. The intermediate representations and textu
503.
by gfodor 4y ago
You should start by building and shipping something even if it is minimal. Starting an open source project with a post calling to form a team and so on is an anti-pattern imo. Good OSS projects always should start with code from one or mayb
504.
by gfodor 4y ago
Yes I was referring to maintaining the training set as-is and just running more than one epoch. I don't think we should expect this to necessarily have the same effect as duplicating data in the training set. I understand it's exp
505.
by gfodor 4y ago
Pricing is too steep to pull the trigger on something so novel and whose utility, quality, and accuracy won't be obvious (or not) until a few tests are done. I'm not sure how to solve this other than to lower them for early adopte
506.
by gfodor 4y ago
My understanding is all of these models trained one epoch. Not knowing the change in loss in these LM’s by doing a second pass seems like a huge blind spot. What’s the best example to point to that could help us understand how likely it is
507.
by gfodor 4y ago
Cool, thanks!
508.
by gfodor 4y ago
Can a physicist comment on the impact of this to the work by Nima Arkani-Hamed et al on the Amplituhedron ( https://en.wikipedia.org/wiki/Amplituhedron )?
509.
by gfodor 4y ago
Yep my reading of everything regarding websockets and durable objects is there will be a per connection metering fee, which sounds like “expensive”, so I opted to go with the much cheaper and likely sustainable pricing of http polling and s
510.
by gfodor 4y ago
Great!

More ›