For perspective, all of the data centers in the US combined use 4% of total electric load.

/checks comments to confirm they’re all arguing over what liberal really means.

Duck Duck Go and Claude concur, fwiw. Except:

If law enforcement directly questions you about a crime, you must be truthful. Lying to federal investigators is itself a crime under 18 U.S.C. § 1001, and similar state laws often exist for state investigations.

Oh, I was really just focusing on video production and entertainment in the next few years. If we’re talking AI’s influence on the information landscape as a whole and humanity in general: I think we’ve discovered how to make a spark- maybe how to gather kindling. We’ll have this fire thing figured out soon, and then who knows what happens. I have no doubt that too much is going to get burned as we learn the dangers and limits of the flame. But civilization awaits us if we survive. I dunno what that means for this technology. Like really, I can imagine so many seemingly equally plausible 2050s that I can’t plant a flag in one. From utopian to dystopian to down right mediocre, I wouldn’t know where to place a bet.

Oh, they’re both quoting James Cagney in White Heat from 1949.

I think it’s from Edgerunners.

So I’ve been producing video professionally for ~25 years, and judicious use of gen ai allows me to do some things that I wouldn’t have the time/resources to do otherwise. As a simple example, Premiere’s generative extend will add a few seconds to the end of (basically) any video clip (basically) seamlessly. Often that’s all the pad I need to improve a cut. The alternatives (re-shoots) are expensive, time-consuming, and approved on a need basis.

Many of the same concerns about the market being flooded with low quality content were raised with the advent of video and again with digital video and again with HD video. The barrier to entry for film is high; for video, it’s virtually non-existent. But I don’t think anyone would claim today that video was a bad idea. AI is in some ways the same kind of democratization of production technology.

I’d be remiss if I didn’t address the ways in which it’s not similar. We can set up a completely automated workflow right now that will quickly generate YouTube “content” and probably make a profit. We could do this before gen AI, but not with such hallucinatory gusto. YouTube is currently being flooded with this crap. But just like people left Twitter (or reddit) when it became overrun by bots, people aren’t going to stick around for your platform full of AI content (at least not until it’s much better).

The IP side of this is mostly funny to me. They’re already talking about a “post-plagiarism” world in academia. I don’t see how copyright survives gen AI at all long-term, frankly. As an artist who saw his first distributed feature film on pirate bay the same day - it just doesn’t bother me. I’ve only ever really gotten paid to do specific work for a client. I don’t expect to get paid for things I make to express myself artistically.

But I hate that I’m shackled to Adobe for a variety of other reasons, and if someone has a good suggestion for an open source alternative to After Effects, I’m all ears.

Edit: No worries. TLDR; YouTube’ll hopefully be forced to alter the toxic algorithms to at least better filter out purely automated shlock.

Algorithms that value engagement over quality are the bigger problem. Stock footage and generative AI are both fine and basically unrelated to this problem.

It’s probably a vision model (like this) with custom instructions that direct it to focus on those factors. It’d be interesting to see the instructions.

To help blind drivers, no. To help AI, yes.

All the companies are gonna merge over the next decade or so, leaving a handful of megacorporations to lord over our cyberpunk dystopia. It’s just easier if all their logos already look the same.

I think it’s more likely a compound sigmoid (don’t Google that). LLMs are composed of distinct technologies working together. As we’ve reached the inflection point of the scaling for one, we’ve pivoted implementations to get back on track. Notably, context windows are no longer an issue. But the most recent pivot came just this week, allowing for a huge jump in performance. There are more promising stepping stones coming into view. Is the exponential curve just a series of sigmoids stacked too close together? In any case, the article’s correct - just adding more compute to the same exact implementation hasn’t enabled scaling exponentially.

There used to be very real hardware reasons that upload had much lower bandwidth. I have no idea if there still are.

I don’t know why it didn’t occur to me before this foggy photo, but that thing sure is cyberpunk dystopian, huh?

Ditto, I was about to start waxing poetic about my bard.

Oh, my friend, how did you come to trade the fiddle for the drum?

The “caravan” of scary South American migrants that Trump fear-mongered about back in '15 (Jesus, has it been a decade?), they were largely subsistence farmers that were forced off of their lands by multi-year droughts that were demonstrably an effect of climate change. In just the last few years, it’s almost comical how dramatically present the effects of climate change are. Whether or not we’re able to admit it to ourselves consciously, I think everyone feels it. We’re about to hit a dozen different asymptotes, and we’re clearly not ready. So, turns out, fear-mongering is a particularly successful strategy in this kind of zeitgeist.

Yes, but most of them involve constant interaction with people that have forgotten how to interact with people.

Cat’s Cradle and Anathem are among my absolute favorites.