They’ve been saying the “bubble is going to burst” for a while now. If they keep saying it eventually they might be right. Then everyone will jump up and down trying to claim they were the ones who rightly predicted it. A quick search by date shows plenty of articles discussing the AI bubble from 2025, 2024, and even a rare few from 2023. Three years of doomsayers waters it down to meaninglessnes even if they‘re eventually right.
it has all the markers, unless the corporations drastically restructure their debt or the product suddenly finds an actual demonstrable use, then it’s going to continue being a bubble. all that requires for it to burst is a large swath of people vested in the bubble to get itchy enough to pull enough money out for the value to be questioned or need to be proved, and AI currently is wildly expensive incuring huge amounts of debt, which is being passed around in a very quick and difficult to account for manner, and has no proven value that can be demonstrated as of yet. to be clear, by demonstrated value I mean you can pull out the paperwork and show where is currently profitable. so here’s the ad revenue, or here’s the government contracts, here the subscription payments, something. they are all operating at a loss and betting their entire companies that those avenues will be profitable eventually, but they aren’t currently and that’s with a VERY heavy hand on the scale, we already have the cost of data center power and water being charged to average people, total failure in providing value to military and police efforts, and people rejecting AI content as acceptable entertainment, it’s all a bust no matter how much they insist, and the application as a digital creative tool is unlikely, it’s jarring and can’t be consistent from one shot to the next causing the exact problems you train to avoid in serial art and animation. digital creative tools generally require a large effort and testing from actual artists in development to be even marginally useful in a production pipeline and AI tried to be useful without any input also with the hope it could simply bypass the people who do the creating, the problem is, it can’t, it can only look at what it sees, fail to identify basic fundamental art principals, and try and remix the art it can look at. look at any AI attempt to recreate cubism, it can’t, it’s totally confused by the concept of incorporating movement and primitive shapes into complex work.
They’ve been saying the “bubble is going to burst” for a while now. If they keep saying it eventually they might be right. Then everyone will jump up and down trying to claim they were the ones who rightly predicted it. A quick search by date shows plenty of articles discussing the AI bubble from 2025, 2024, and even a rare few from 2023. Three years of doomsayers waters it down to meaninglessnes even if they‘re eventually right.
it has all the markers, unless the corporations drastically restructure their debt or the product suddenly finds an actual demonstrable use, then it’s going to continue being a bubble. all that requires for it to burst is a large swath of people vested in the bubble to get itchy enough to pull enough money out for the value to be questioned or need to be proved, and AI currently is wildly expensive incuring huge amounts of debt, which is being passed around in a very quick and difficult to account for manner, and has no proven value that can be demonstrated as of yet. to be clear, by demonstrated value I mean you can pull out the paperwork and show where is currently profitable. so here’s the ad revenue, or here’s the government contracts, here the subscription payments, something. they are all operating at a loss and betting their entire companies that those avenues will be profitable eventually, but they aren’t currently and that’s with a VERY heavy hand on the scale, we already have the cost of data center power and water being charged to average people, total failure in providing value to military and police efforts, and people rejecting AI content as acceptable entertainment, it’s all a bust no matter how much they insist, and the application as a digital creative tool is unlikely, it’s jarring and can’t be consistent from one shot to the next causing the exact problems you train to avoid in serial art and animation. digital creative tools generally require a large effort and testing from actual artists in development to be even marginally useful in a production pipeline and AI tried to be useful without any input also with the hope it could simply bypass the people who do the creating, the problem is, it can’t, it can only look at what it sees, fail to identify basic fundamental art principals, and try and remix the art it can look at. look at any AI attempt to recreate cubism, it can’t, it’s totally confused by the concept of incorporating movement and primitive shapes into complex work.
Id like to think that people are more aware of the circumstances that lead up to a bubble forming and popping.