Raw LLM Responses

Hmmm, am not exactly qualified super qualified but I can give a bit of perspective from a CS background on how the industry seems to want to expand…. First, the current big target goals of AI are: Being sold as a service for big business, in form of data analysis, decision making support and automation of certain aspects of the tech and non-tech industries. In the case of the public the goal is to reach “AI agent” capabilities, which essentially means having the AI be able to control and manipulate the environment it runs in, providing assistants on steroids essentially. All the cringy AI things they try to force you to buy now are mostly to get you to buy the phone out of AI hype. The AI isn’t the product yet, but it’s close to becoming the product as requirements drop. Second I don’t think it’s going anywhere… mostly because even if it where to fail on a large business scale (I can see that happening) the models that already exist and you can personally run locally aren’t nearly as expensive as I thought they would be. You can, as an individual with a high end pc run your own versions of stable diffusion and deepseek. Higher quality models of those two require decent sums of money but that’s quickly going down… fast. The current 5090 + 64gb of RAM can run the middle-class deepseek model Unless AI optimisation halts and hardware stops improving I don’t see how in a few years it won’t be relatively easy to run your own version to train yourself Ofc that’s not profitable for big business but they don’t get a say on that part after making the models public I don’t think this is dying because contrary to NFTS, AI does seem useful for people, very useful. Even if big tech drops out of the race and the models stop being improved I think AI will stick around