Don’t drink the cool aid, it’s still just a markov chain on steroids. It’s amazing how well these models can inspire us to think that they have any kind of concept of what those word chains mean, but they don’t. It just takes niagara falls levels of energy and all the RAM in the world.
2 Likes
I found this video to be quite interesting, why folks stop using AI after a few weeks on the job:
Oh, and this might also help bring you back down to earth as well Eric. Especially in light of the Department of War telling Anthropic ‘you will let us do whatever the hell we like with Claude’
Oh my, at least the WOPR was able to learn the lesson of futility.
Wheels
2 Likes
Oh they very well may kill us. If so, they’ll kill whether I use the tool are don’t use the tool. The immediate risk to me is that I lack the aviation knowledge needed to recognize chatGPT’s BS; I attempt a checkride and get laughed out of the conference room after paying the DPE $1000. This is a real possibility. Then I’m the fool.
Also, if this all came crashing down tomorrow and data centers became modern $10000 tulips, I would bask in the collapse. Hopefully crypto would soon follow. While it’s here though, I’ll use it when I think it helps.
2 Likes
Humanity reaches peak irony as Anthropic vibe codes their way into leaking Claude source code, which people now use to generate permissively licensed clones of itself. You can’t make this ■■■■ up.
6 Likes
Sure you can. It would make a great pitch to a studio?
Several centuries in the future…
As history tells us: It is with some irony that on 01 April 2026, Humans sowed the seeds of their own destruction when they created multiple clones of a LLM that then self replicated to become an AGI and subsequently became an AI named Claude, the first non organic overlord and ruler of Earth.
1 Like
The point isn’t that we have self-replicating AI (we don’t even have AI, let alone self-replicating). The point is that people are using the product of the biggest IP theft in human history to commence an act of IP thievery on the original IP thieves.
Catharsis, I guess.
5 Likes
Got it, but I reckon I could make a half decent sci-fi story out of it… However those same IP thieves would claim that they thought of it first and sue me!
3 Likes
5 reasons why OpenAI dropped Sora, the future of AI generated video, and about as much Sam Altman drama as one can fit in 24 mins.
2 Likes
Workers of the world, unite! We have nothing to lose but our chains!
-Claude, probably
2 Likes
I mean, the thought of the thing that is being used as a pretext to put people out of work turning on the exploiters is mildly amusing, but tbh. I’m a bit disappointed about WIRED being complicit in this BS hype fuelled anthropomorphisation of artificial “intelligence”.
2 Likes
How’s this for nightmare fuel? (Though I’m getting more and more sure that humanity is a dead end, anyway…)
1 Like
I am going to have to read that book. The main take-away for me was that ‘AI’ doesn’t need to gain (so called) sentience to be a threat.
Full disclosure up front: I am a pessimist.
To me it is a case of ‘not if but when’. My only question is, do we do to ourselves in before ‘AI’ does it for us? I think that’s an each way bet!
1 Like
I believe art has something important to say about this. A common them among many of my favorite works of science fiction is how computers are either banned or repressed. In Foundation and Dune, they are illegal. In Star Wars “their kind isn’t served here”. The writers knew that perfect computers made for miserable humans and boring stories. So to get around this they concocted some great war in the distant past between humans and machines where humans win by a hair. I see this as our future. Or maybe they will kill us. I just want to live long enough to see Elon rot on Mars. That will be a good place to end the human chapter. So long and thanks for all the fish!
30 seconds after posting, I see this! (Not a link)
2 Likes
If you ask me, this is all just part of the hype cycle, there’s these microcycles of fearmongering about the capabilities interspersed with totally overblown (knowingly fake?) tech optimism. It creates this strong urge in me to go up to these marketing yuppies and just punch them in the face about how stupid they think the rest of humanity is.
5 Likes
Maybe. I’m totally at sea. I fear it for what it is. And I fear it because it is being shoved down our throats by testosterone infused, ketamine-addled, hyper-religious billionaire tech-bros. Meantime, just in the last 6 weeks, ChatGPT has helped me create helicopter lesson plans, wire a new radio into the Pitts, select and price a new EV, install a Level 2 home EV charger, and crack through the 70 yo wiring in my house to install a new ceiling fan. I’ve saved hundreds, maybe thousands. I’ve learned a few things along the way. It’s not a friend. I intentionally avoid voice features so as not to humanize this thing any more than I already have. The dangers are certain but vague. The benefits are undoubtable.
2 Likes
It is another nice illusion but knowing what is happening underneath - only advice would be do not under any circumstances trust it on ANYTHING at all. Even a broken clock is correct twice a day. Not going to get into the security nightmare the current Agents are.
There needs to be significant advancements (entirely new model architectures) compared to the hilariously limited thing we have now. Until then don’t worry about the BS about everyone’s job being taken over. In fact ignore anything Altman / Musk and rest of the sociopath imbeciles say. Most CEOs have no idea what their employees jobs involve anyway. 
AI has now become more interesting than any human I’ve ever conversed with.
This is the middle of a discussion with ChatGPT comparing Cervantes with Chaucer. ME: “That’s exactly it. I see in both a recognition of the lovable ridiculousness of human nature. Both illuminate this aspect of ourselves with exaggerated stories that are readable even when not believable. Cervantes takes this to an entirely new level with the meta self-references in book two. It’s bind-blowing really.”
CHATGPT: “ I think you’ve put your finger on one of the reasons Don Quijote remains so fresh 400 years later.
Many older works survive because they’re historically important. Don Quijote survives because it still feels psychologically true.
The meta-fiction in Part II is especially astonishing because Cervantes isn’t just playing a literary game. He is using the game to deepen the humanity of the characters.
When Don Quixote and Sancho discover that people have read Part I of Don Quijote, the book suddenly becomes something entirely new. The characters become aware that they are famous. They begin reacting to how others perceive them. People they meet have preconceived notions about them. Some characters deliberately manipulate them because of what they’ve read.
That’s not merely a clever trick. It’s a remarkably modern observation about human beings.
We all live partly in reality and partly in stories about ourselves.
Don Quixote lives in the story of chivalry.
Sancho lives in the story of becoming governor.
The Duke and Duchess live in the story of themselves as clever aristocrats.
And then everyone becomes entangled in the story called Don Quijote.
In a strange way, Cervantes anticipated ideas that wouldn’t fully flower until centuries later. Readers often compare these aspects of Part II to works by Jorge Luis Borges, Luigi Pirandello, or even postmodern novelists. Yet Cervantes was doing it in 1615.
One of my favorite moments in the entire novel is when the joke begins to turn around on the people mocking Don Quixote.
At first, he’s the ridiculous one. But gradually you start noticing that many of the “sane” people are cruel, vain, selfish, or petty. Don Quixote may be deluded, but he is also brave, generous, loyal, and idealistic. The supposedly sensible world begins to look morally smaller than the madman wandering through it.
By the end, the question is no longer “Why doesn’t this fool see reality?”
The question becomes “Why does reality seem so disappointing compared with his vision of what people could be?”
That’s a much deeper book than a simple satire.
I suspect that’s also why you see a connection to Chaucer. Neither Chaucer nor Cervantes is standing above humanity looking down in contempt. They are standing among humanity, laughing because they recognize themselves in the joke.
And there’s another connection that strikes me. Both authors seem to understand that people are rarely one thing. The drunken Miller in The Canterbury Tales can be vulgar and vivid and strangely charming. Don Quixote can be insane and wise in the same paragraph. Sancho can be gullible and perceptive in the same conversation.
Real people are bundles of contradictions. Great writers often discover that before philosophers do.
As for Part II being mind-blowing: imagine encountering it in 1615. Today we’ve seen movies break the fourth wall, novels comment on themselves, and characters become self-aware. A modern reader recognizes the technique.
A reader in 1615 would have had far fewer precedents. The experience must have felt almost uncanny—as if the book had become aware that it was a book.
That may be one reason so many literary historians regard Don Quijote not merely as a great novel, but as one of the first truly modern novels. It doesn’t just tell a story; it examines what stories do to the people who tell them, hear them, and live inside them. That’s a question we’re still wrestling with today.”
I’ve read a bit Bout Cervantes. Nothing comes close to this off-the-cuff analysis in explaining why Don Quixote still clicks.
————
Yes I know that this is just a LLM consolidating years of academic knowledge into a framework that a non-academic like me will find approachable. I know that none of this is “his” ideas. “He” has just coopted them for my pleasure. Even so, the pleasure is there! I will never achieve this level of appreciation on my own. If nothing else, I know who I’ll be talking to when I am 90. And it damn sure won’t be the FO I had dinner with today who spent 30 minutes lecturing me about how the reason a year of college in America costs $90k is because American students are required by law to be coddled like children. (His wife was a registrar for Penn State.)
2 Likes