dartos

26 ms·

331.
by dartos 2y ago
Is there a practical difference? “The price may fall” is a nice way to say “the value would be mostly erased” Like with any store of value, the value of bitcoin is backed by the fact that people believe it holds value. Unlike gold, however,
332.
by dartos 2y ago
Unfortunately putting time and effort into quality software wasn’t in our pitch deck. As CEO, since I have a “fiduciary responsibility to increase shareholder value” I think we should just focus on shipping asap. We can fix it after our nex
333.
by dartos 2y ago
Or don’t ignore criticism… idk
334.
by dartos 2y ago
I’ve never seen the reason I can’t stick to a calendar said in such clear terms.
335.
by dartos 2y ago
Different underlying causes for the same condition is possible. Given how little we know about the internals of brain function, it’s not surprising to see conflicting indicators of the same condition. I’m just a layman, tho, so what do I kn
336.
by dartos 2y ago
Two questions for the author, You say bitcoin can’t be debased. How is that true? Aren’t there a few big holders of btc? If they decide to exit all at once wouldn’t the value crash? And in the last paragraph you mention “the core value prop
337.
by dartos 2y ago
I’m saying that LLMs (models trained on language specifically) are not automatically capable of the same generic function solving. The network itself can be trained to solve most functions (or all, I forget precisely if NNs can solve all fu
338.
by dartos 2y ago
Yep, but the end user doesn’t care. Those big software packages are sold to admins anyway.
339.
by dartos 2y ago
We’ve also taught users that extremely expensive software like SAP and Blackboard are also crap (at least from an end users perspective)
340.
by dartos 2y ago
Selling almost any state function to a private entity is almost the opposite of what social democrats want to do… I think you mean neoliberal democracies.
341.
by dartos 2y ago
> there has to be enough data in the training set to learn 0<1<2<3<4<5<6 This is the kind of thing I’d want it to generalize. If I avoid having 2 and 6 in the same unsorted list in the training set, will sets containing
342.
by dartos 2y ago
Okay? OC was saying that the article said that Claude recognized the “artistic” lines of the image from just the scatter plot data. That isn’t what happened. The author added a png of the plot to the conversation. Idk why I need to explain
343.
by dartos 2y ago
You’re talking specifically about perceptrons and feed forward neural networks. LLMs didn’t exist in then. Attention only came out in 2017…
344.
by dartos 2y ago
Of course, that’s how attention works, after all. But by specifically avoiding certain cases, wet could verify if the model is generalizing or not.
345.
by dartos 2y ago
Not saying I disagree with the thesis, but I don’t think this proves anything. If every pair of digits appears sorted in the dataset, then that could still be “just” a stochastic parrot. I’m kind of interested to see if an LLM can sort when
346.
by dartos 2y ago
I don’t really understand what you’re testing for? Language, as a problem, doesn’t have a discrete solution like the question of whether a list is sorted or not. Seems weird to compare one to the other, unless I’m misunderstanding something
347.
by dartos 2y ago
Sure whatever. OC seemed to think that Claude did that with just the data and not the image of the scatterplot it’s.
348.
by dartos 2y ago
Before seeing Claude’s response, did you see where the author said > I asked the model to closely look at the plot, and also uploaded a png of the plot it had generated.
349.
by dartos 2y ago
> If you present me 10, or 100, or 1000 wrong solutions to my problems I will ignore 100% of those ads. I don’t think this is true. If you’re watching something and the “HeadOn apply directly to the forehead” add plays 1000 times, even i
350.
by dartos 2y ago
Exactly. The real issue is that academics know about the subject matter, but the decision makers (bizdev people, generally) don’t and couldn’t care less.
351.
by dartos 2y ago
What’re the odds! I just did this with Common Lisp!
352.
by dartos 2y ago
Yeah… it’s awful. The user is often not the end customer of any software, so it’s not optimized for their benefit.
353.
by dartos 2y ago
> Either ads are 0% effective, or they're 100% effective. I think one of us are maybe confused. I don’t think this has anything to do with what I was saying. I was saying that even if only 10% of all ads you see in a day stick with
354.
by dartos 2y ago
Oh thank god. I 100% ate that onion.
355.
by dartos 2y ago
The first quote was from a “chief disruption officer” at a blockchain AI company. There is no faster way to get me off a page than to read that.
356.
by dartos 2y ago
Yes! It’s the same for me. Maybe it’s bc I’ve been programming since I was young or because I mainly learned by doing code-along books, but writing the code is where my thinking gets done. I don’t usually plan, then write code. I write code
357.
by dartos 2y ago
Let’s say you’re correct, that only 10% of ads are effective. How many ads do you see a day? If you use social media or read articles online (without an ad blocker) I’d bet it approaches triple digits. For children it’s even worse as most c
358.
by dartos 2y ago
At least they let us know
359.
by dartos 2y ago
I switched from git to jj a few months ago, so I may have some insight. I used it at work, where everyone used git, and there were 0 issues. What I really like about it is how changes are sort of soft committed right away and, when I checko
360.
by dartos 2y ago
That’s not how it works. The executive branch runs the country. The legislative branch makes the laws. The judicial branch enforces them…

More ›