Warning: Don’t eat before watching it.
1 Like
I love the Jank. It’s the best part about it.
2 Likes
They’re doing the same in the movie industry. The people hyping Gen AI up clearly have a stake in it and it’s so painfully obvious.
4 Likes
When I constantly read and see people saying AI is taking the jobs of creatives and AI is going to be making movies and games, I’ve ironically yet to see any evidence that it can achieve anything remotely resembling a finished product and not just churn out an unusable tech demo.
The hype surrounding AI is very forced. We’ve all used it by now for various things, i.e. it’s open source. And the results are just not something anyone could charge money for.
5 Likes
The concern is the churn. Like if you’re Disney and all you’re doing is rereading old ideas, what’s the difference? Like I saw Tron:Ares, aside from the nine inch nails track, gen AI could’ve written that entire movie. Hollywood as is isn’t incentivizing originality, and AI excels at poorly regurgitating old ideas.
AI is most dangerous with ignorant groups looking for an easy magic answer to all their problems. If you want an actual story of it stealing jobs I can give you one. I’m working to teach English as a second language and the school I work at, doesn’t do a very good job teaching English. The country I’m in doesn’t. They don’t really fundamentally understand it, and as a result they don’t properly teach the fundamentals. Case in point the school I’m at doesn’t teach English writing at all. And in this environment they’ve started to introduce AI writing as a shortcut and magic fix. That’s the real danger of generative AI. Its a shoddy tool that looks flashy. We don’t actually have any really good uses for it (as of now), but it’s very persuasive for those that think it can break down non-existent barriers in the arts & humanities. And then there’s the investors all in on the AI rocket ship to trillion dollar stock infinity (nevermind the fuel running out).
So yeah this stuff looks terrible to all of us and we’ll call it out on its BS and we won’t buy into it, but there are people who are out there buying into it a LOT. All while gen Ai perpetuates the worst of the internet and continues to be subpar or uncanny or something fun for bored college students to mess around with. It isn’t taking over, but it’s convinced certain people that it is. And those people are going all in for it (and unfortunately they don’t live in a bubble; how we each interact with it will affect us all).
4 Likes
Do people not understand what “going to be” means? As in… the future? We are basically in the stone age of AI. It’s impressive that something like this demo is even possible. (Although it’s probably just an AI-generated video?)
Also, I’m pretty sure that this Twitter post is meant to be sarcastic?
Regardless, the PC Gamer “news” article and the comments are funnier to me than any AI demo. And I doubt that any of those people are aware of the irony of calling AI dumb. I had to double-check the date to make sure it’s not an old post from April 1.
2 Likes
You could make an entire E3 press conference filled with fake game trailers. The AI would probably struggle with making a realistic Phil Spencer, but since it’s all fake anyway, you could probably have people like the Rock Tony Hawk, and Bill Gates present. Sounds kind of fun actually, over the years I’ve had plenty of game ideas. Grand Theft Auto Texas, Xbox FPS Maker, Xbox Kart, Xbox Fighting… Having Master Chief as a character in Tony Hawks DLC. I’d create an announcement trailer for the next Gen Xbox as well, previously I made a AI logo for “Xbox Hybrid” that looked legit.
It’s a tool like anything else, hyperbole of AI will be normal life in 5 years
1 Like
I’d say there’s more likely to be a pushback based on the tool’s inherent limitations. It’s prompt based. Maybe people who don’t work in visual art don’t really know the difference but artists will find these tools extremely limited.
I tried ComfyUI and found myself basically fighting against its limitations.
That looks horrible and the sooner the AI bubble bursts, the better off everyone will be.
4 Likes
The way something like this should be viewed is what it might be able to produce in the next few years. And how much money that might make.
The AI bubble isn’t going to burst. It will shrink down a bit based on reasonable expectations, while it grows over the years as the tech becomes more capable.
If you’ve never seen the Gartner hype lifecycle, I think we’re coming off the peak of inflated expectations into the trough of disillusionment, before it normalizes into the later stages. Maybe we’re already heading out of the disillusionment part. But it’s here to stay in a big way.
The thoughts I’m hearing from tech investment leaders in my circle are things like “it’s like an intern who just joined the company,” or “it helps automate well defined processes with predictable input and output,” and “expect up to 80% of a good enough starting point with good prompting.”
There are pockets where it is helping across business functions. And even in the last few months the improvements are apparent, as well as a better fit between expectations and practical applications.
6 Likes
“AI bubble burst” will IMO arrive quite organically due to its inherent limitations. It won’t suddenly explode, it’ll simply lose steam due to lack of mass adoption. They can’t force this through. We’re already seeing tech firms struggling to find ways to make people and companies adopt this tech as the revolution many with a vested interest in lining their pockets with this thing want it to be. Copilot? lol, no thanks. AI chat bot are far too inconsistent to be a reliable source. There’s too many unfixable errors (hallucinations) there for it to ever take off. And who decided people wanted a middle man between their Google search and the results? And don’t get me started on suggestions game consoles could be fitted with one of these. It’s got those Xbox One reveal vibes where Yusuf Mehdi tried to sell the thing off the back of… buying movie tickets for Star Trek whilst watching Star Trek. The entire premise of chatting to a bot is basically wildly overrated.
Then there’s the image and video generators which rely on prompts. Artists can’t use these (controlnet is a joke in ComfyUI) and the whole shebang feels like a variation on a video game like Dreams. It’s just an unreliable toy which alternates between ‘meh okay that’s kind cool but what’s the point?’ results and ‘that’s borked’ abominations.
And in the artistic field the key issue remains: no one is paying money to consume AI generated products.
4 Likes
The issue being is that’s nothing normal automation tools or machine learning couldn’t already do, without the dangers of hallucinating.
At the moment the costs are similar, but to make their money back the price is going to have to rise (and is already eye watering at the corporate level, but then so are most cloud-based automation systems) and that may cause people to reconsider.
I’ve seen more than a few companies and CEOs stake their corporate lives on automation or AI replacing much of their workforce and be much cheaper - volume processing ones particularly - and are now screwed as it turns out anything further than their current basic automation will cost far more than paying basic rate for staff.
I agree that there probably won’t be a huge burst, but that it’ll wind back then ramp up again once we finally start to near AGI or at least AI that hallucinates less and has better contextual awareness.
However the downturn may coincide with the private credit industry starting to pop, with banks winding in their lending and so less money being available for investment - enough that it might start to look like a burst if the stock market starts to fall too…
2 Likes
Unfortunately they are through advertising - no one would pay money to watch an AI generated film, but the AI slop that overwhelms social media (often pushing false stories in weird robotic voices) is far too popular with idiots and so earns a lot of advertising revenue.
The amount of times I have to point out to older relatives that a video or image is AI slop and to stop sharing it…
I agree no one is paying to watch AI stuff, but I do fear as some less reputable companies realise they can get more eyeballs and revenue from mass-producing cheap slop than well-crafted content
2 Likes
Let’s hope so, actually, because what I’m seeing more and more is execs and bean counters with absolutely no business overseeing creative fields absolutely misread the market and skew their interpretations in a way which suits the speculators and shareholders. So if gullible social media addicts are giving them a false sense of ‘win’, I’m all for it.
The day they’re releasing AI generated video games after laying off the actual talent (at $80 of course because hey ho no one said this had to be good for the consumer right? Margins are king) we’ll see who blinks first. Same goes for those useless bots.
Small anecdote: I seldom use AI bots at all but when I do… oh boy. I was googling Alien (the movie) a few days ago and Copilot automatically popped up with its summary and proceeded to inform me that Brett was killed by the Alien whilst investigating the location on the ship Kane died in. I mean… people want to use this stuff to research cancer treatment? Ha.
3 Likes
IT spending in Mid-market and enterprise is skewing towards more AI projects, not less.
I wouldn’t read too much into the highly publicized fumbles of orgs that went too far, too fast. That’s just reckless leadership.
There are a lot of level headed people who are taking it in at a more reasonable pace, with reasonable expectations on how to test and learn where it can be applied. And quick to draw a line where things need to stay business as usual until the tech improves. The common theme though among tech investors is that continuous testing and learning on this front, across the business, is expected.
The examples I know of aren’t well-publicised ones, they’re companies I’m close to and know the plans of - too often, the dream of automation and AI doing everything has been sold as “just on the horizon” and idiots / people without enough knowledge fall for it, even when those around them are trying to warn them it’s not as close as people think.
In one case, a firm has incredibly low profit margins on high-volume revenue, but is hoping to be acquired by venture capital by suggesting its workforce can be replaced by automation and AI very soon (and that this will offset the huge investment in AI they require) - in reality, they’ve been promising that for well over a decade now and have automated all that they can, anything further the costs are coming out two or three times their employee costs and there’s simply not enough profit in the industry they’re in to cover that.
Instead, it’s likely to end up with huge cost cutting impacting their staff before eventual winding up / sale to a third party in the sector, with their employees screwed while the bosses walk away with a fat payday (don’t get me started on predatory / late stage capitalism).
On a public note, there’s Uber which has had a ridiculously high value for years due to the fact they’re always promised they’re going to replace their expensive drivers with self-driving cars and the profit margins will grow fat - instead, that target is forever over the horizon and they’ve instead raised prices, attacked any employee protections in order to keep costs low and generally been a shitty company.
Yes there are many CEOs or CTOs who listen to their IT departments and understand the market, and know the limits of AI as it stands and where it could go in the near future - but there’s still way too many investors and captains of industry who are just buying the hype and just aren’t aware how far we still are from AGI and that betting the house on it in the short term is not just silly, it’s potentially disastrous for their employees and the economy in general…
1 Like
Then those leaders who navigate this change the wrong way, with negative business performance outcomes, will naturally fall out of leadership positions. It’s unfortunate that they will take a lot of well intentioned and talented people with them, but that’s how these things work. The leadership with clear eyes and steady hands will likely come out on top.
1 Like