Pages
▼
11 December 2025
Video of the day -- from fakery to disaster
For years now the "AI" hypesters have promised godlike super-intelligence and a radical transformation of the very nature of work. All they've delivered has been novelty gimmick programs that generate boring fake pictures and videos, and chatbots and search-engine "summaries" that are constantly wrong about everything. "AI" in the workplace turns out to be more trouble than it's worth and is too unreliable to be entrusted with anything important. And this unbroken record of pitiful failure is costing hundreds of billions of dollars with no end in sight.
It's the vast sums of money involved that transform the "AI" phenomenon from a mere embarrassment into a serious danger. In this video, Adam Conover gives the clearest and bluntest explanation I've seen of the failure of "AI" and the economic catastrophe it portends for the near future.
Looking around the blogosphere and seeing people endlessly discussing Trump and elections and so forth, I sometimes have the sensation of overhearing a group of people debating who should sit where around a picnic table on a beach, oblivious to the huge tsunami heading straight toward them. This video explains the tsunami.
5 comments:
Please be on-topic and read the comments policy. Spam, trolls, and fight-pickers will be deleted. If you don't have a Blogspot account and aren't sure how to comment, please see here. Fair warning: anything supporting transgender ideology, or negative toward Brexit, or in favor of a military draft or compulsory national service, will be deleted. I am not obligated to provide a platform for views I find morally abhorrent.
No comments advocating violence against any specific identifiable individual, even jokingly.
Please be considerate -- no political or politics-tinged comments on non-political posts, and no performative cynicism (or cynicism in general).
As I have recently posted, it can be useful in minor ways for photo editing, but I am really disliking AI in general.
ReplyDeleteI am commenting based upon my personal history of ~ 25 years of quietly designing, testing, deploying and analyzing knowledge systems against real world problems.
ReplyDeleteI don’t disagree with most of what’s been said here but with a critical caveat - 99.9% percent of what people think of (and comment about) as AI are really LLM, Large Language Model, or generative AI models. That’s where all the hype is and that’s where all that cash has been spent. AI is a much broader area
Our own group developed an early version of LLM many years ago and continue to use it in very focused applications. This version remains useful because it was always anchored in the definition of language NOT the statistical use of words or phrases in content. Since LLM is a rhetorical approach to language it suffers from the same repetitive use issues as humans in social media environments.
AI, like human intelligence, is not one thing. The human brain is an amalgamation of a number of anatomical and electro-chemical processes that work in concert. True AI will be much the same. It will be a suite of technologies with specialized roles within a domain. That’s the approach we’ve taken and the results are highly encouraging. (In fact, the problems w LLMs discussed here have been a boon to us in helping people understand why our systems are designed they way they are.)
The key to true AI is pretty much the same as the key to successful human critical thinking - anchoring decision-making in iterative processes that have a transparent and unbroken chain of evidence to true knowledge. Applying relentless and unbiased assessment of the outcomes of the decisions and continuous process improvement. It works.
We are starting to see a shift at the academic level toward the knowledge basis approach we are using, driven in large part by the problems with LLM, and are hopeful that we will see they first deployments of our systems at scale.
As for the hype surrounding AGI (artificial general intelligence) - most of it is just that. But not all. 2026 may be an extremely interesting year on that front.
Anvil: Very few people seem to like it (except for the novelty value of chatbots, and the fact that some people get hooked on them), but it's being pushed on us relentlessly.
ReplyDeletePliny: I can't speak to what you've been working on, but the point here is the economic collapse which is going to result from the failure of "AI" to achieve any of the amazing things the hypesters claim it will. And there's no sign of "AI" by any definition being able to achieve such things in the foreseeable future. If there were technology available that could produce true artificial intelligence or make all these revolutionary changes in the workplace, companies like OpenAI would be deploying it so they could finally make some real revenue. Instead, we're stuck on this trajectory toward a massive stock market crash and subsequent collapse of much of the economy.
GREAT video! I love Adam Conover! His shows are always fantastic!
ReplyDeleteGlad you liked it. For myself, I had not seen his material before.
ReplyDelete