darawk

48 ms·

451.
by darawk 5y ago
It doesn't seem like you read what I said. I'm not defending "tough on crime laws". And I specifically said we should be softer on non-violent crime. So i'm not sure what it is you're responding to.
452.
by darawk 5y ago
It's almost as though the framing of "tougher" vs "softer" is a ridiculous over-simplification of the matter. We should be tougher on violent crime and softer on non-violent crime. EDIT: I want to amend this to say
453.
by darawk 5y ago
> Either way, like someone else in the comments said, these past 2 years have shown us that laypeople are incapable of following active science in real time and drawing reasoned conclusions. Frankly, this is probably true of everyone who
454.
by darawk 5y ago
Sure. But the rate you will buy it at will change.
455.
by darawk 5y ago
It's really not debatable at all. What is debatable is how much printing it takes. It is not debatable that there exists an amount of printing that will cause inflation.
456.
by darawk 5y ago
No, they can default even when it is denominated in their own currency, which is what Russia did. https://en.wikipedia.org/wiki/1998_Russian_financial_crisis Most countries just choose not to do this. They weren't
457.
by darawk 5y ago
> "The United States can pay any debt it has because we can always print money to do that. So there is zero probability of default." Alan Greenspan While this is partially true: 1. Sovereign nations absolutely can and do defaul
458.
by darawk 5y ago
You say that, but you are probably pretty sure I am doing the thing you call "cognition", and I am pretty sure you are. Despite the fact that the only evidence we have for that is the essentially meaningless strings of text we&#x2
459.
by darawk 5y ago
I'm not assuming they share anything. I'm saying that the contrary has not been established. Whether or not I believe they are in the same category is distinct from the question of whether or not it has been proven that they a
460.
by darawk 5y ago
That's exactly the point, though. The converse is also presented without evidence. We have no evidence that human cognition and GPT-3 cognition are distinct in some fundamental way. All we really know is that we are better at it than G
461.
by darawk 5y ago
To everyone saying "Well it's just obviously a security", you may well be right. But securities law and the Howey test are actually pretty complicated. If the SEC has a rationale for why this product is a security, that'
462.
by darawk 5y ago
In this context, his job is to assess the legality of it before they create the product. Companies don't generally want to invest in building something they aren't legally allowed to market.
463.
by darawk 5y ago
If that were the case, then the SEC should say so. The fact that they are unwilling to suggests that there is more to it than you think.
464.
by darawk 5y ago
Totally agree. I was excited by this journal's existence, and then disappointed when everything in the first issue was philosophical in nature, and not scientific. The 'papers' in there right now are basically just unusually
465.
by darawk 5y ago
You shouldn't let the politics of the situation cloud the obvious. If they didn't have guns, it would not have played out the way it did. That doesn't mean that what they did was good, or that how the government handled it wa
466.
by darawk 5y ago
Regular guns are sufficiently potent weapons to give the government trouble. Whatever else you might think of him and his stunt, Ammon Bundy proved that.
467.
by darawk 5y ago
The judicial system has basically nothing to do with monetary policy.
468.
by darawk 5y ago
SF seems like a pretty serious location to me. Granted, its still manned, but it makes sense that they'd start out manned in any new environment out of an abundance of caution. It seems likely that we could see a switch to unmanned in
469.
by darawk 5y ago
99% of the market does not even know about telemetry and spyware. And then about 90% of the 1% who does does not actually care. Smart TVs are popular because they offer a far better user experience for the average person. It really is that
470.
by darawk 5y ago
Honestly though, do we? Why? It almost certainly wasn't man made and intentional. That's the only case we really need to know the origin story for, and I don't think anyone serious thinks that is plausible. Otherwise, the wor
471.
by darawk 5y ago
Don't conflate "meaningful" with "predictive". The attributes are indeed predictive. The intention of the model is to make accurate predictions optimally given the data its fed. If you give it data that contains pre
472.
by darawk 5y ago
Totally agree. I found the total lack of equations, data, and experiments in the paper this thread is about extremely telling in this regard as well. Lots of vague assertions, zero rigor. I think all the points being made in this paper evap
473.
by darawk 5y ago
So I think you and I agree on all the things you just said. My point is really just that, linguistically, I don't think it makes sense to describe the models as being "biased" or "discriminatory" here. Statistical l
474.
by darawk 5y ago
I think your example is obscuring the issue here a little bit, because you didn't really describe a targeted learning problem: > It is trivial for algorithms to introduce novel racism: consider processing text analysing minority cau
475.
by darawk 5y ago
> But the paper is talking about a different kind of bias, say, that a data set of loans has thousands of data points of people born in the first neighborhood and tens of data points of people born in the second neighborhood. Even if you
476.
by darawk 5y ago
> Models are not neutral. All models encode inbuilt priors reflecting some inherent bias. In fact, absent that bias a model would have little to no generalizability beyond its training set! Yes, but those structural priors have nothing t
477.
by darawk 5y ago
Ya, the overloading of 'bias' is confusing. > The "algorithm" doesnt capture the semantics of the data, and so introduces moral bias regardless of any statistical bias within the dataset. > Eg., consider training a
478.
by darawk 5y ago
I don't think "it's a data problem" is inherently used to hand-wave away the problem. I think it's used to locate the problem in the area where its most readily addressed. Designing algorithms to debias your data is
479.
by darawk 5y ago
> A surprisingly sticky belief is that a machine learning model merely reflects existing algorithmic bias in the dataset and does not itself contribute to harm. Why, despite clear evidence to the contrary, does the myth of the impartial
480.
by darawk 5y ago
I mean, I guess that could happen. But it's not clear why we should be upset about something that might happen in the future. This software doesn't really get us any closer or further from that happening. If Amazon wanted to keylo

More ›