> Or maybe the next great story teller to create something meaningful that touches people in ways that change their mental model of the world.
How many pieces of AI art can you actually remember? I mean, call to mind in the same detail as you can remember a photograph?
I think AI generated imagery is fundamentally compromised somehow in this regard: something subliminally uncanny, no matter how realistic, makes them harder to recall.
For this reason I personally doubt AI generated art will ever have a profound effect on people. Because it really seems to lack the mechanism.
Well it is that but I think it is also that it is, at some informational level, fundamentally incoherent and unreal. It looks fine, but it is not anything. It has no intent in the art strokes (which I think always shows in geometry), it has no reality in the lighting of a photo.
It may be, I concede, that I see more AI-generated photos than other art types (AI generated photo fraud is a serious issue in a corner of the web I frequent) but I tend to find that I literally can't remember what they look like long after I see them.
Same exercise, focussing on faces specifically:
- try to visualise Taylor Swift's face. Or that of Rachel Weisz or Ming-Na Wen, or Sarah Silverman, or Alfre Woodard.
- now try to visualise the face of Tilly Norwood.
Obviously if you don't know who any of these people are, you can't do this exercise (which is why I included Taylor Swift). And if you don't know what Tilly Norwood is, you can't do this exercise.
But if you've seen a lot of content about Tilly Norwood, can you visualise the face in the same way? Is it memorable? It is not.
It is my contention that these images actually have something very undefinable missing, that my brain needs to find them worth memorising. I have seen many "AI models" now and I can't remember any of "their" faces.
> comments in this submission is just straight up guessing and assuming whatever guesses they make are correct.
True of every comment thread on HN.
Absorbing the thoughts of other humans on any topic you have deep knowledge of makes you see that all coverage of EVERY topic is subtly incorrect / poor / has an agenda.
For four great years in my career I ran game servers for an Australian ISP. I really enjoyed tossing up servers for new HL2 / UT / Quakeworld mods and seeing what picked up a community, chipping in on a cyberpunk HL2 mod in the same period.
I have such rose coloured glasses of that time.
Looking back, most of the dedicated server software felt like it was just tossed over the wall. Some of the stuff we used to have to do to get things running happily on headless linux servers was very hacky. Others simply HAD to run on windows hosts.
I feel like the entire industry died as games became "live" services.
Nearly 2 decades later when my kids got into Minecraft, I stumbled into the hosted MC server world and was just amazed by the size of the industry around it.
It was a real "arrrh this is where that same spirit ended up" moment.
And now of course there's huge servers funded by getting kids into gambling and pay to win.... Gross.
FWIW, I use Fidesmo. Oversimplified, it allows you to copy your credit card's NFC chip into an accessory you wear. I use a ring but there are other options like bracelets or watch bands. No batteries, no devices, no wireless connectivity. It works anywhere an NFC card works, which here in Switzerland is more or less everywhere.
It requires that the card issuer support Fidesmo though. Many here do but I'm not sure what it's like elsewhere.
That's not how those NFC cards work. They are payment middlemen. They are full cards on their own and just pass on every charge to your other card. Just like Google Pay.
Sounds very likely. Perhaps if you are sufficiently big you could also get a small kickback from someone like VISA? Operational expenses must be fairly low.
The way I described it was oversimplified. Technically, it's more like your credit card issuer issues a new card with the same number and installs it on the chip in the accessory.
To be able to do it, you have to authenticate with your card issuer in a mobile app, similar to how you might when setting up Android Pay or Apple Pay. The mobile app then uses your phone as a bridge between the issuer and the NFC chip in the accessory so the relevant data can be written in a secure way.
NFC payments via Google Wallet running on my Pixel Watch 3 connected to a phone running GrapheneOS works just fine. I use this regularly. (It doesn't require Google Wallet to be installed on the phone.)
At least one of my cards required Google Play Services to have the location permission when initially adding the card though.
I've had a similar approach. My kids computers are setup next to mine and I keep an eye on what they're playing.
I've instigated a purchase wait period of at least 3 days. Very often they themselves realize that the thing that they wanted to spend their pocket money on was a brief desire.
I was super proud when I heard my son say "meh, this is pay to win" as quitting a random roblox game he was trying out.
They don't have to have those. Depending on your definition of "kids", most people on HN I imagine are not giving their kids phones, laptops, or tablets at young ages (maybe less than ~13?). And if they do, I imagine the devices are somewhat locked down and monitored.
I think the more technologically literate a person is, the more wary they are of unfettered access to it for children. Hence, preferring a stationary desktop where use can be supervised.
I agree desktops are best, and they are what my kids started with, but there is a lot of pressure to give kids phones.
For example, where I live, the cheapest (monthly) bus tickets require an app, so kids need a smartphone to get to school (or their parents have to pay a lot more for daily tickets).
There is a lot of social pressure on the kids too. There are lots of activities that have either moved online or are organised online. Lots of ways to get left out.
It's an open-source framework for building intelligent, task-driven bots in Minecraft Java Edition powered by large language models (LLMs) such as GPT-4, Gemini, and many others. Designed for research and creative automation, these bots can connect to Minecraft worlds, perceive and act on their environment, undertake custom tasks like resource gathering or building, and even collaborate in multi-agent scenarios.
Perhaps allowing them to find the positives in new technologies.
Or, you know... Slop.
An open mind doesn't hurt though.
reply