Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
elicash
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
11 ms
·
211.
▲
by
elicash
1y ago
And then there's more, via 404: “Since RedHat doesn't want to answer to us,” the hackers wrote in a channel on Telegram viewed by 404 Media, suggesting they have attempted to contact Red Hat. [...] “We have given them too much tim
212.
▲
by
elicash
1y ago
"The hackers stated that they attempted to contact Red Hat with an extortion demand but received no response other than a templated reply instructing them to submit a vulnerability report to their security team." Just hilarious
213.
▲
by
elicash
1y ago
This is like saying it's a bad thing that phone cameras are getting so good because expensive cameras were keeping out the garbage films. Expensive cameras are still better. And yet, it's good that people who were never going to b
214.
▲
by
elicash
1y ago
1. Hallucinations have been reduced. There are, in fact, reasoning models. The "ifs" are not as theoretical as you imply. But the reason I raise the question on what happens if the acting is of high quality is because it's im
215.
▲
by
elicash
1y ago
I think two things are being communicated with "actor" 1. Unlike traditional animation, these AI characters are being "directed" in a sense that's unlike traditional drawing frame by frame. 2. The broader question o
216.
▲
by
elicash
1y ago
> AI slop acting So if the technical challenges are solved and the AI acting is quite good, that changes your view of it? > I just don't understand how this would work unless the entire movie was just CGI... if real actors are su
217.
▲
by
elicash
1y ago
I already have a better chess engine at different skill levels for 1-player mode. For two human players, I plan to start with sending a link to a friend given there won't be enough random players on the website to find one in real-time
218.
▲
by
elicash
1y ago
Working on a chess / poker hybrid. There was "choker" back in the day, which I actually never heard about since I wasn't into chess back then. But (1) there was no web version, and (2) it had a specific gameplay that see
219.
▲
by
elicash
1y ago
Getting rid of editors happened at the same time as shorter deadlines. I'm sympathetic.
220.
▲
by
elicash
1y ago
My approach has been to "yolo" my way through the first time, yes in a somewhat lazy and careless manner, get a working version, and then build a second time more thoughtfully.
221.
▲
by
elicash
1y ago
I feel like there should be a term that's like "MVP," but instead of a "minimum viable," it's a FULL FEATURE SET but just super sloppy and not really usable for production. FSV = full slop version
222.
▲
by
elicash
1y ago
They go up to $4,500 (solid gold), but they start at $81 on their website.
223.
▲
by
elicash
1y ago
My understanding was that life expectancy has been declining for like a while, not merely slowing in gains -- first because of opioids and then because of COVID. Edit: In the U.S. that is.
224.
▲
by
elicash
1y ago
> Notwithstanding context This is actually quite important! Especially when you're not talking about a single word/name but a group of several words/names.
225.
▲
by
elicash
1y ago
Results here are bad, obviously, but it'll be interesting when LLMs can not just identify fonts but unredact pieces of documents in places where just a few words are removed by analyzing the length of redaction, combos of letters that
226.
▲
by
elicash
1y ago
There are already small newsletter businesses that offer basic products.
227.
▲
by
elicash
1y ago
This post linked to a different article (former Bloomberg reporter) that said it expected this round to be at $750 million ($50 million above the previous valuation). And so yes, that's significantly less than it ended up being, $1.1 b
228.
▲
by
elicash
1y ago
I was on a trip and two nights in a ROW the battery went from 100% to 0% overnight while closed and on standby. It turned out I was just leaving it too close to the split A/C unit at the airbnb.
229.
▲
by
elicash
1y ago
Lots of us were bad at this even before AI.
230.
▲
by
elicash
1y ago
I agree with this. But also, part of what made the original study interesting is the DISCREPANCY between perceived vs actual productivity, as opposed to which method "won." It implies that even if some tasks are better with AI tha
231.
▲
by
elicash
1y ago
I'm building things at a level of complexity I wouldn't have even attempted without AI. This piece, however, only focuses on time spent on a task that could be done both ways. Even there, it falls short. Let's assume this stu
232.
▲
by
elicash
1y ago
They are trying to communicate that disabling AI is a product feature to some customers. I think it's reasonable to, for example, have a field that is labeled "Do Not Track".
233.
▲
by
elicash
1y ago
Yes, in the video on OpenAI's website, they reference a product that Altman is testing at home created by Ive. No, we can't see it yet. And there's not much description, either. Just that it's the "coolest technolog
234.
▲
by
elicash
1y ago
Not sure if it counts for what you meant by hardware, since it's more art piece than a practical thing given the $60k price, but he created a record player: https://www.linn.co.uk/us/turntables/lp12-50
235.
▲
by
elicash
1y ago
Please write this up somewhere!
236.
▲
by
elicash
2y ago
Irony that the person who assumed the worst of the article author is offended by bad faith accusations. But I do believe you that you just didn't read it.
237.
▲
by
elicash
2y ago
He got a legal threat when he was 19 and claimed to be no longer hacking people's paypals. We don't actually know that he was a minor when he stopped. I think it's fine to say that things done even at 19 or younger shouldn&#x
238.
▲
by
elicash
2y ago
You are quoting only a few words out of the article to create a false impression. Here's another short excerpt, an alleged quote: > I no longer hack into Paypals, gain root access into other peoples computer (sic), or exploit online
239.
▲
by
elicash
2y ago
Folks are correct this is dangerous. But you could imagine a world where batteries were required to be built in a way that this type of tinkering of individual cells and matching them was safer. If it could be done, would certainly would be
240.
▲
by
elicash
2y ago
What did the prompt look like for Leonard.ai? I'm curious if you said 2d animation style for both or just for chatgpt. Edit: Your second version of chatgpt doesn't say photorealistic. Can you share the Leonard.ai prompt?
More ›