cs702

49 ms·

691.
by cs702 2y ago
> Coppola has been trying to make this movie for more than 40 years, during which the project has gone through innumerable rewrites, delays and false starts. It exists now only because he sold part of his successful winery estate to fin
692.
by cs702 2y ago
Explicit ≠ literal. The things they mention/demo -- response times, multiple domains, inflection and tone, etc. -- are those that make it feel "magical."
693.
by cs702 2y ago
Yes, the announcement explicitly states that much of the effort for this release was focused on things that make it feel magical (response times, multiple domains, etc.), not on moving the needle on quantifiable practical performance. For f
694.
by cs702 2y ago
The usual critics will quickly point out that LLMs like GPT-4o still have a lot of failure modes and suffer from issues that remain unresolved. They will point out that we're reaping diminishing returns from Transformers. They will que
695.
by cs702 2y ago
Well, that really sucks . Thanks But No Thanks.
696.
by cs702 2y ago
Market cap is now close to ~$11B, for a company that over the last 12 months burned ~$240M in cash. Revenues declined by ~11%. Net cash on hand is ~$600M. It's hard to justify purchasing the shares at the current price based on the com
697.
by cs702 2y ago
If the price starts dropping, the fearful start dumping, and volume jumps. Sellers pushing the price down beget more sellers pushing the price down. Hedge funds will try to ride it up and down.
698.
by cs702 2y ago
Market cap is now close to ~$11B, for a company that over the last 12 months burned ~$240M in cash. Revenues declined by ~11%. Net cash on hand is ~$600M. It's hard to justify purchasing the shares at the current price based on the com
699.
by cs702 2y ago
That's a really big number. I'd guess it represents ~100% of customers who have given information to Dell. The hackers say they signed up as a Dell partner, instantly giving them access to a partner web portal, from which they har
700.
by cs702 2y ago
Hahahaha. They will learn a lesson that can be learned in no other way.[a] Regardless, all this entrepreneurial activity is great news for the country! --- [a] With apologies to Mark Twain: https://www.goodreads.com/quotes
701.
by cs702 2y ago
By many accounts, Tesla FSD has improved rapidly, now feels "very human," and will soon have greater adoption than any other self-driving software.[a] Speaking from personal experience, I've tested the most recent version of
702.
by cs702 2y ago
This is 100% right. The original purpose of scientific "journals" and "meetings," as those words themselves indicate, was to enable scientists to share new ideas and findings more efficiently with each other. There was
703.
Tesla FSD free trial conversion rate is ~2% so far(old.reddit.com)
3 points by cs702 2y ago|0 comments
704.
by cs702 2y ago
Alright. I've added you preprint to my reading list, so I can take a closer look at this.
705.
by cs702 2y ago
Can you show the world this can be made to work for, say, a toy benchmark like MNIST classification? --- To be 100% clear: My question about practical application today is orthogonal to the question about whether this research is worth pu
706.
by cs702 2y ago
When I ask if this has been tested, I meant as an ANN on conventional benchmarks. Sorry if that wasn't clear. There are a lot of ideas that are clever and seem promising... but fail to perform well on such benchmarks. Is there a github
707.
by cs702 2y ago
Has this been tested? Doesn't look like it.
708.
by cs702 2y ago
Over the past couple of weeks, I've come across quite a bit of anecdotal evidence suggesting that a lot of people in the US -- but especially those earning below median income -- badly need to earn more to make ends meet. Food, energ
709.
by cs702 2y ago
I agree. It seems the edge cases that trip these systems are becoming so rare and convoluted that the return on this kind of research is approaching zero. At some point we just need to collect data from real-world tests.
710.
by cs702 2y ago
You're right. It's one of the few kind-of-uplifting Dark Mirror episodes. > Having an automation that filters by your preferences and solves the whole “swipe fatigue” issue seems brilliant. I doubt it will work, because users
711.
by cs702 2y ago
And she said it with a straight face . This new, ahem, "feature" reminds me of the Dark Mirror episode "Hang the DJ."[a] Reality is getting as weird as, or even weirder than, dystopian fiction. I don't even know if
712.
by cs702 2y ago
Given how bad human drivers already are, I wouldn't be surprised if there are lots of ways to fool them that don't fool the machines. Speaking from experience, having used Tesla's "Supervised FSD" v12.3.6, which is
713.
by cs702 2y ago
https://www.google.com/search?q=confusing+road+signs&udm=2 Illusions aren't even necessary with human drivers, many of whom are shockingly bad -- look around next time you're on the road.
714.
by cs702 2y ago
These guys had to work really, really hard to find a way to fool the machine vision of recent self-driving software! My take: Maybe the machine vision of recent self-driving software has become harder to fool than human vision? Human vision
715.
by cs702 2y ago
> The world is still trending toward a dystopian hellscape, but the dystopian hellscape is not a barren planet where nothing grows and we've stripped everything bare, it's a dystopian hellscape of everybody glued to their devi
716.
by cs702 2y ago
I'm not sure "peak baby" is a good thing, but agree with you that the other three peaks are encouraging. Arguments by smart people like Sabine Hossenfelder have led me to question the conventional wisdom on population growth.
717.
by cs702 2y ago
Reports of the slowdown of scientific progress are wildly exaggerated.[a] Scientific -- and mathematical -- progress continues in fits and starts, as always, thanks to individuals who labor in relative obscurity, year after year. Journalist
718.
by cs702 2y ago
I think the grandparent comment is about AI research, driven by the quest for AGI. Incremental improvement of proven approaches, driven by profit motive, will surely continue regardless of whether there is an AI winter or not.
719.
by cs702 2y ago
If there are no breakthroughs and funding dries up in the near future, it will feel to many like going off a precipice at high speed . Only the poor souls who survive the fall, at the very bottom of the precipice, will get to experience th
720.
by cs702 2y ago
This deserves to be on the front page. The authors ask whether image-to-text and text-to-image models (like CLIP and Stable Diffusion) are truly capable of zero-shot generalization. To answer the question, the authors compile a list of 4000

More ›