This isn’t meant for those who are already wise to this AI bubble (and Wall Streets complicity in this rigged game that is going on). Unsurprisingly, there’s a huge number of channels on YouTube now that are taking AI generated summaries, fronting it with an AI generated person (or in other cases, simply audio with AI generated visuals), and just farming subs, views, and engagement metrics.
Here are just a few channels as examples (which I am not directly linking to): Doctor Laura (which at least has a disclaimer on the profile disclaimer and on the videos), Hardware Alert, Hardware Details, and HD Knows (you will see this same AI avatar of this asian guy on other channels talking about lifehack or health topics). Note there are many more (some labeled as AI, many others not — having a healthy dose of skepticism does help navigate this stuff).
Recommendation: always perform your own due diligence. We are reaching the “trust nothing” phase with everything.
There is another channel called Project Nightwatch where the actual channel owner (to his credit) made a video to explain why he created the channel (admitted many of them use AI generated visuals though none of which were labeled as such). This included a few that blew up after the M7.7 quake off of the Sanriku coast because that particular video was 1h47m of looping information which had the hallmarks of being generated by AI prompts (he says he researches all of this stuff because he has an interest in this subject matter).
Note: I used my other account BTW so as to not pollute my main one; I pulled that one video to watch it for myself to see why it had over 540K views and could understand why it could draw people in during the day the quake occurred and the subsequent week after. But it’s 1.25 hours too long where it becomes obvious after a certain point that it’s farming for engagement metrics (on the total watch hours).
That becomes more evident when most of the other videos on that channel are similarly 1.5-2 hours long. Reading the comments, some people eventually did catch on caught on to what was going where the views on recent videos have dropped off. I have a hard time believing it’s not AI generated given the sheer amount of info that’s required to fill 1.5 hours (even a short posting like this which is at best just a few minutes of reading time, takes a while to draft up and I use no AI for help with formulating a post).
With regards to content that is both fronted by an AI generated avatar and spitting out info from an AI prompt, YouTube really should place the “AI” notice up top (that is if the channel owner even bothered to mark their content as being AI generated or modified) versus buried in the description. Users of these platforms also need to take up a more defensive posture as these tools get better.
BTW, this is what this AI bubble is being partially fueled by; not just the slop of videos of cats cooking/doing people things, of dogs wearing masks and scaring other pets, of people falling into these holes in the ground, etc., but of informational content being presented by a pleasant person that is simply AI generated. These YouTube channels have a vested interest in getting into the algorithm in order to get those subscribers and views in order to generate revenue to justify the money they are spending for AI to help create their content.
But that is all burning through a lot of energy and those costs are going to be increasing for these companies given this energy crisis that has yet to fully slap the global economy in the face. IMHO, we are at the Webvan and Pets.com point of the dot com bubble in terms of where the AI bubble is at. I experienced that bubble because I was living and working in Silicon Valley at the time and was also heavily invested in many internet companies where I made out on some and was decimated in others (and all I can say is this current bubble is far worse).
