The solution to global warming is "deploy solar, wind, hydro, and storage en masse, and improve city infrastructure so that more people can walk, bike, and take public transportation rather than using their car". All AI will do is tell us that, but that's not the answer people want to hear.
frezik
joined 1 year ago
Make everything faster. Space that isn't used for caching data is space that's wasted.
This isn't necessarily about apps that run on your desktop or phone. Most code in the world runs on servers, and the use cases are different.
We've already hit a perceived user experience limit. The perception of responsiveness in blind tests between SATA and NVMe SSDs isn't always apparent--people sometimes say the SATA drive is faster--even though the speed difference on paper is substantial.
IMO, programmers haven't exploited the possibilities of extremely fast mass storage yet. The orders of magnitude difference in speed isn't fully realized. It's not just faster, it's faster in a way that requires new approaches. Unlike multicore CPUs over a decade ago, this change in thinking has gone relatively unnoticed by programmers.
GP was wrong about tapes, but plenty of these systems use hard drives already. They can use specialized drives that are cheap and have slow write speeds, because streaming video is a constant rate per second. They also don't record unless there's movement. The network is also a limiting factor.
I don't think SSDs solve any problem, here.
This is per chip. A given NVMe drive usually contains multiple chips.
Is that what has happened to the storage market historically?
That's likely the point where spinning platters die in the marketplace.
Right now, spinning platters are around $12/tb. SSDs are around $75. Exact numbers fluctuate with features and market changes, but those are the ballpark. Cut in half, SSDs will be $38/tb, and then $19 in the next halving. Spinning platters aren't likely to see the same level of reduction in that time period; they're a mature technology.
I think once they reach double the price per tb, we'll see a major collapse of the hard drive market. My thinking is that there's a lot of four drive RAID 10 systems out there. With SSDs, those can be two drive RAID 1, and will still be faster. With half the drives, they can be twice the price and work out the same.
If we're talking about what Moore originally formulated, then the law isn't just about transistors. He actually claimed that the cost per integrated component is cutting in half every x months. The exact value of x was tweaked over the years, but we settled on 18 months.
If we were just talking about transistor count, the industry has kept up. When we bring price into the mix, then we're about an order of magnitude behind where we "should" be.
When he wrote it, the first integrated circuit had only been invented about 6 years prior. He was working from only 6 years of data and figured the price per integrated component would continue to drop for another decade. It's remarkable that it lasted as long as it did, and I wish we could find a way to be happy with that. We've done amazing things with the ICs we have, and probably haven't found everything we can do with them. If gate sizes hit a limit, so what? We'll still think of new ways to apply the technology.
I hate how this phrase has been abused so much. There's nothing particularly extraordinary here--we're not talking about bigfoot or aliens--and the whole point of a documentary like this is to lay out evidence.
A single individual is the most likely way to keep a secret compartmentalized.
A huge number of apps these days are web sites compiled into an app, and it shows. For example, an app should be able to remember your address and payment information without signing into an account, yet so many don't. Almost like they want to force you into signing up. Why might that be?
Just give me a mobile web page if you're going to do that shit.
I won't believe AGI is here unless it has an existential crisis within 100ms of being turned on.