It’s not just the playing, even the buying can be a chore, as you’ll have to dig through dozens of different versions, DLC, and season passes to figure out what you are even buying, most of the time the actual online shop doesn’t even tell you, you have to search around forums to figure out what you get. Starting one of those Ultimate Edition that includes everything also means spending 5min clicking though dozens of “You just bought DLC” notifications.

Getting late into a game series is also always “fun”, as you can’t even tell what is a prequel, sequel, spin-off or whatever, as most content no longer puts a number in the title. That’s another trip to Wikipedia, as I have yet to see any online shop providing that information.

Needless to say, I stick mostly with older or indie games. I can’t stand how every modern game needs to have skill trees, collectives, level ups and hundred different weapons that all look and feel the same.

That said, chores can also be quite subjective. The Riddler trophies in the Batman Arkham games can certainly be seen as chore when you just want to reach the end fast, collecting them takes around three times as long as the main game. I however found them to be the best part of those games, as they are very old school and based in exploration and puzzles, as opposed to just running from cutscene to cutscene. They give the player a lot of agency and freedom that is missing in the main plot.

Not great, but also not really Adobe’s fault. Journalists using them are the problem. Back when the fakenews of Israel hitting that hospital with 500 dead went around there were plenty of news article that just had regular stock images from other completely unrelated bombings in the articles, which did nothing more than misguide the reader. That’s the kind of stuff that really shouldn’t be acceptable, but happens all to often.

The media needs much better standards when it comes to photos (e.g. include GPS coordinates and time so we can verify and cross check it easier). Just plastering stock images in articles, AI generated or not, is rarely helpful and often misleading.

Sounds like somebody accidentally bought some germicidal bulbs instead of regular black light. Not the first time this has happened. See BigClive’s videos on the topic.

I think it’s worth pointing out that this really isn’t an “AI issue”, but a “visual effects issue”. Actors have been getting replaced with visual effects for 30 years, just recently Nicolas Cage filmed scenes for Flash, only to be completely replaced with a different CGI scene. Sometimes you don’t even need CGI, e.g. in The Mandalorian you have a stuntman under the mask a lot of the time with the actor just doing voice over. AI adds another layer of flexibility to the visual effects, but when you look at something like Gemini Man, the tech is basically here to do completely realistic face replacement even with just classic CGI.

The “identity theft” thing is also somewhat disingenuous when you consider the work of stuntmen, which get little to no credit compared to the main actor, despite often doing a lot of the actual work. The actors identity is basically just a marketing tool. Though even that has been fading in value long before AI, with modern movies being based around franchises, not actors or directors.

Anyway, given how good AI video already starts to look, I really don’t mind AI entering the field. It allows smaller teams with less budget to produce something that doesn’t have to hide behind the $300 million movies in terms of effects. Will still take a while until we get there, but it’s inevitable.

I am sure the problem will magically fix itself when we switch the cameras off.

Anybody remember Euphoria? Also seen in that canceled Indiana Jones game.

The point of the AI doomsday warning is that we are at a point where we still can think about those problems. Fast forward a couple of years and AI will be integrated as deeply into our society as electricity. There no longer will be an option to just switch it off when everything we use depends on it.

Anyway, that aside, I feel the whole scaremongering about bots and propaganda is a bit misguided. Yes, those are real issues that can and will happen. But it’s neither AIs fault nor can it be brought under control by regulating AI. To fix that we have to uplift our journalistic standards by like a lot. At the moment nobody even tries to provide useful information, nobody tells you their sources, nobody tells you GPS coordinates or the time when something happens, nobody even provides plain old links, it’s all just “hearsay” and “trust me bro”.

Just a few days ago the worlds mainstream press was reporting how Israel bombed a hospital and killed 500 people, all while the hospital was still standing with no bomb crater in sight. Some article even included stock photography of other completely unrelated bombings in the past. The whole event was trivial for everybody to verify, yet nobody did. Worse yet, some even doubled down on their initial misreporting days later (e.g. Channel4).

So yeah, with mainstream journalists being less trustworthy than random guy on Twitter, we do have a problem. But that problem is not with AI. If anything, I hope that AI can help us out of this misery by automating the fact checking and sourcing of information. At the moment even seemingly basic task like “where did this photo came from?” require far to much manual work.

Also everything is just streaming. You can’t actually buy DRM-free movies (important for things like VR headsets or when without Internet), often you can’t even buy anything at all, it’s all just the library of the streaming service were content comes and goes at random.

• Another World
• The Longest Journey
• EF2000

Just have the LLM output verifiable scripts instead of manipulating the data directly itself and have version control for the data so the AI can undo changes. All pretty doable, though maybe tricky to get into old apps.

Yes, but whoever overcomes those problems will be the next Microsoft/Apple/Google (or get rich by getting bough by either of them). I think a large paradigm shift in how we do computing is unavoidable, LLMs are way to powerful to just be left as chatbots.

I think we might end up with the Microsoft/Apple/Google situation all over again. While it’s easy to build an AI, having to jump between AIs for each and every task is no fun. I think the one that wins the golden goose is the one that manages to build a complete OS with AI at it’s core, i.e. instead of Unix shell, you just have a ChatGPT-like thing sitting there that it can interact with all your data and other software in a save and reliable manner. Basically the computer from StarTrek were you just tell it what you want and it figures out how to get it.

That others can spin up their own LLM won’t help here, as whoever gets to be the default AI that pops up when you switch on your computer will be the one that has the control and can reek the benefits.

but it doesn’t take a decade.

GTA2 to GTA3, two years. GTAIV to GTAV, five years, GTAV to GTAVI 10 years and still waiting. See a pattern?

When you want to accurately simulate and animate every doorknob in the game it just takes a hell of a lot longer than putting a few pixel on a texture and calling it a day. And yes, there is in argument to be made that that level of detail doesn’t actually add to the gaming experience in a meaningful way, but that’s something a whole lot of AAA games struggle with and why we are seeing so many sequels and so little new things. StarCitizen just happens to be a new thing developed from scratch, including the studio itself.

Star Citizen has literally the highest budget of any game in history, by a factor of 2x. The second biggest one is Cyberpunk, which also has been in development since 2013. Gamedev has just spiraled a bit out of control and takes ridiculous amounts of time these days.

The only reason you see yearly releases of some other games is because they have multiple teams working in parallel and are just updating an already existing game engine. Doing stuff from scratch is a whole different thing.

That’s not quite how it works. The pixels are just the first layer. Those get broken down into edges. The edges get broken down into shape. The shapes get broken down into features like eyes, noses, etc. Those get broken down into faces. And so on. It’s hierarchical feature detection. Which also happens to be what the human brain does.

The actual “drawing” the AI does is quite a bit different however. The diffusion works by starting with random noise and then gradually denoising it until an image emerges. While humans can approach painting that way, it’s rather rarely done so.

The goal with the X rename was to make a western equivalent of WeChat, that’s a Chinese app for chatting, social media and payment. While X was the name of the payment company Elon had before it got turned into Paypal.

Crux is, outside of the name change, we have yet to see “X” gain any more features that would bring it closer to WeChat, it’s still just a broken version of Twitter.

“New snake oil to give artists a false sense of security” - The last of these tools I tried had absolutely zero effect on the AI, which is not exactly surprising given that there are hundreds of different ways to make use of image data as well as lots of completely different models. You’ll never cover that all with some pixel twisting.

that was already mitigated and addressed

Yeah, after a year. Sorry, but I don’t take lightly to companies that are stealing screenshot of my browser and than act like it’s no big deal.

without any hint of what they would be.

Have you not been paying attention over the last few years? Mozilla’s numerous missteps ain’t exactly a secret. Here is a little list:

• https://digdeeper.club/articles/mozilla.xhtml

Brave. It ain’t perfect, but I actually like that it comes with Adblock, IPFS and Tor support out of the box. Gives you a fully functioning browser out of the box without having to mess with tons of plugins.

If you want something more minimalist, Librewolf might be worth a look.

Pocket is Mozilla’s bookmark/sync pay-cloud-service. Comes with Firefox by default and can’t be easily removed. From a company that claims to care about privacy I would expect a self-hosted local-first approach for such problems, not a cloud service.