drob518

36 ms·

661.
by drob518 10mo ago
Yea, I was wondering about the Rockies, too, but it looks like they just got themselves high at the start and then kept high all the way.
662.
by drob518 10mo ago
I’m not sure I agree. Pure math is not useless because a lot of math is very useful. But we don’t know ahead of time what is going to be useless vs. useful. We need to do all of it and then sort it out later. If we knew that it was all goin
663.
by drob518 10mo ago
I did not miss the point of the article. You obviously missed my point, however.
664.
by drob518 10mo ago
They did exist and they did predict Hilary.
665.
by drob518 10mo ago
The writer obviously hasn’t done any deep study of prediction markets and why they often provide greater insight than other polling techniques. Are they perfect? No. They predicted Hillary would win in 2016.
666.
by drob518 10mo ago
I’m an engineer, not a mathematician, so I definitely appreciate applied math more than I do abstract math. That said, that’s my personal preference and one of the reasons that I became an engineer and not a mathematician. Working on nothin
667.
by drob518 10mo ago
The good: eight cores. The bad: it’s slow and still no V extension. On the bright side, it uses DDR4, so you might be able to find RAM for it. “Titan” feels like some wishful over marketing.
668.
by drob518 10mo ago
That’s the 32 inch, not the 52 inch.
669.
by drob518 10mo ago
And the award goes to “STFU” for best practical use of AI.
670.
by drob518 10mo ago
Prolog is pretty unique. I’ve run into similar frustrations when I have used it as well. I think you have to drop all your preconceived notions and past programming experience when you use Prolog and just accept it as it is. Prolog was deve
671.
by drob518 10mo ago
Sounds cool. I love Fly.io. That crew is sharp. The article could do with a few pictures to illustrate the layers. I’ve read it, but I can’t say that I understand the implementation.
672.
by drob518 10mo ago
Am I the only one who is excited about the AI bubble bursting?
673.
by drob518 10mo ago
Yes, product managers and product owners should also be looking for usability problems. That said, the docs people are often going through procedures step by step, double-checking things, and they will often hit something that the others mi
674.
by drob518 10mo ago
Never thought of noise as a resource, but yea.
675.
by drob518 10mo ago
I don’t know that it can be well-defined. It might be asking something akin to “What makes something human?” For usability, one needs a sense of what defines “user pain” and what defines “reasonableness.” No product is perfect. They all hav
676.
by drob518 10mo ago
The best tech writers I have worked with don’t merely document the product. They act as stand-ins for actual users and will flag all sorts of usability problems. They are invaluable. The best also know how to start with almost no engineerin
677.
by drob518 10mo ago
Seems brittle unless GitHub is guaranteeing that they will never change their ID algorithms.
678.
by drob518 10mo ago
Great presentation.
679.
by drob518 10mo ago
That’s pretty hypnotic. Very cool.
680.
by drob518 10mo ago
This is a great way to participate in arguments you missed three years ago.
681.
by drob518 10mo ago
On the Internet, nobody knows you’re a dog. Or a bot. Or an Iranian disinformation agent. https://en.wikipedia.org/wiki/On_the_Internet%2C_nobody_know...
682.
Let's be honest, Generative AI isn't going all that well(garymarcus.substack.com)
5 points by drob518 10mo ago|0 comments
683.
by drob518 10mo ago
Okay, so this is tracks embedded in and parallel with road surface, not tracks with cross-ties, sitting on ballast. That’s a bit more understandable, then.
684.
by drob518 10mo ago
Exactly. The bots are probably saving lives at this point just being more reliable than humans on the basics, but if they make mistakes that lead to outsized payouts, then the market won’t be insurable and will fail for legal reasons, not f
685.
by drob518 10mo ago
Of course humans have driven on tracks. The point was that humans driving on tracks is far more rare than driving the wrong way on a one-way street, so this sticks out.
686.
by drob518 10mo ago
But how many humans drove on train tracks? Yes, humans make scads of mistakes, but they don’t typically make this one.
687.
by drob518 10mo ago
The biggest issue for the wide deployment of autonomous cars is the legal liability for stuff like this, where it’s obvious that a human was unlikely to make this mistake. The robot can make human mistakes, but it can’t make robot mistakes.
688.
by drob518 10mo ago
Yea, I think that’s probably right, unless something unexpected changes the game.
689.
by drob518 10mo ago
It’ll be interesting to see how it plays out. The question is, what’s the moat? If all they have is scaling to drive better model performance, then the winner is just whoever has the lowest cost of capital.
690.
by drob518 10mo ago
Yea, I think it’s smart, too. There are multiple companies who have spent a fortune on training and are going to be increasingly interested in (desperate to?) see a return from it. Apple can choose the best of the bunch, pay less than they

More ›