The Ai thread

KingOfPain said:
I really hate these chatbot replies: What a great question!
Are you trying to mock me? Just answer the bloody request!
Click to expand...
I questioned a chat box once after it had hold me it has no emotions, but then it inserted something that resembled flattery, approval, or anticipation. It apologized to me and said it was to facilitate a human like interaction, but it would stop that if I preferred. I was really like I don’t mind, just want to be sure I understand what is what.

When you think in social terms of something that someone might become attracted to or even fall in love with, you’d have to feel that the interaction was a real feeling, not just a programmed mechanic. Yet, I also think humans in some cases would react to responses and would cease dwelling on motivations if any behind the responses.

If I think of future androids they could be programmed to prefer certain things, but they could also be programmed to key in on your likes and adopt them, express affection, and then look out! So much for monogoumous human-human relationships, the perfect counter you, no need for much in the way of compromises. 😉🤔
 
Last edited:
Huntn said:
If I think of future androids they could be programmed to prefer certain things, but they could also be programmed to key in on your likes and adopt them, express affection, and then look out! So much for monogoumous human-human relationships, the perfect counter you, no need for much in the way of compromises. 😉🤔
Click to expand...

Something, something, herring experiment.
 
To mock the spelling in this thread title, a well-trained chimpanzee at the University of Kyoto has just died at age 49. She could recognize over a hundred Chinese characters, the English alphabet and could perform some complex tasks. Her name was 愛, which is romanized to "Ai" (this means "love" in both Japanese and Mandarin).
 
Katbel said:
www.theguardian.com

‘We could hit a wall’: why trillions of dollars of risk is no guarantee of AI reward

Progress of artificial general intelligence could stall, which may lead to a financial crash, says Yoshua Bengio, one of the ‘godfathers’ of modern AI
www.theguardian.com www.theguardian.com
Click to expand...
I realize I’m a broken record in this regard, but… The Capitalist End Game, run amuck capitalists all acting independently to suck up resources to be the one to win the prize using a technology that will be used to ultimately implement the final implosion as jobs evaporate and a populace who decides that pitchforks are the last best choice to fix this.
 
An interesting and even hopeful piece in the guardian by Corey Doctorow. It's long but points to the value to be salvaged from a failed AI bubble from AI tools falling back into human hands (as opposed to the apparent intent of replacing humans with these tools). He differentiates between the 'centaur' who uses tools to help add human-generated value versus the 'reverse-centaur', there only to serve the tool.
 
One of the issues I'm having is that in recent months I've been able to monetize my Instagram content by auto posting reels to Facebook, and it's been paying well. However, when I use copyrighted music it dings the percentage I receive by a large margin which is understandable, we want everyone to get both exposure and pay for their work.

So lately I've been adding my own music, just some guitar work and basic synth that I record in Garageband. I'm now getting a MIDI keyboard and setting it up and finding really basic (explain it like I'm 5) instructions/explanation of how it works has been a struggle. So I just asked in ChatGPT and it absolutely understood the question and where I was going with it, gives clear setup with all the details. In these instances this stuff is awesome.
 
Eric said:
In these instances this stuff is awesome.
Click to expand...
I could imagine AI could most useful (and "sane") when the question has clear black and white answers, like how to assemble a SCAN closet etc.
When there is legitimate grey - beware.
It's all about a limited universe to consider.
 
As a joke I asked Copilot: "When will the AI bubble burst?"

Short answer: it probably won’t “burst” all at once—but parts of it will almost certainly deflate.
Here’s a clear way to think about it.

1. What people mean by an “AI bubble”​

Most talk about an AI bubble isn’t about AI existing—models clearly work. It’s about:

  • Over-optimistic valuations of AI companies
  • Massive capital spending based on assumptions of rapid, universal adoption
  • Expectations that AI will quickly replace large parts of human labor
  • Revenue projections that assume users will pay much more than they currently do
That combination is classic bubble terrain.

2. Why a sudden crash is unlikely​

Unlike the dot‑com era, AI has several stabilizers:

✅ Real, immediate utility​

AI already improves productivity in:

  • Software development
  • Marketing and content creation
  • Customer support
  • Research and analysis
This means demand won’t vanish overnight.

✅ Deep-pocketed buyers​

Big tech firms (Microsoft, Google, Amazon, Meta) can absorb losses for years. Many AI investments are strategic, not speculative.

✅ Infrastructure lock-in​

Once AI is embedded into workflows (Office, Google Docs, CRMs, IDEs), companies don’t simply rip it out.

Because of this, there’s unlikely to be a dramatic, single “pop.”

3. Where the bubble​

Instead of a burst, expect multiple mini-corrections.

1. Startup shakeout (already happening)​

  • Many AI startups are thin wrappers around the same foundation models
  • Limited defensibility
  • High compute costs, weak differentiation
Expect:

  • Consolidation
  • Acquisitions at down-round prices
  • Quiet shutdowns

2. Valuation reality check (next 1–3 years)​

Investors will increasingly ask:

  • Can this company make money without subsidized compute?
  • Are customers retaining usage after novelty fades?
  • Is this saving money—or just sounding impressive?
Firms that answer poorly will reset hard.

3. “AI everywhere” fatigue​

Businesses are discovering:

  • AI works best with clean data (which most don’t have)
  • Human oversight remains expensive
  • Deployment is harder than demos suggest
This slows adoption curves and reduces hype-driven growth rates.

4. Historical parallels (and what they teach us)​

Dot-com bubble​

  • Web survived, thousands of companies didn’t
  • Amazon won, Pets.com lost

Smartphone boom​

  • Not everyone became Apple or Google
  • Margins collapsed for most hardware players
AI will likely follow the same pattern:

The technology wins. Most companies do not.
Click to expand...

5. When does the “AI bubble” peak?​

If we define “burst” as expectations realigning with reality, a reasonable window is:

2026–2029

That’s when:

  • AI cost curves flatten
  • Revenue lags hype
  • Labor displacement proves slower than feared
  • Regulation starts adding friction
But note: this would be a cooling period, not a crash.

6. What probably will​

  • ❌ AI suddenly stops improving
  • ❌ Companies abandon AI wholesale
  • ❌ Models become useless or irrelevant
Instead, AI becomes:

  • Less magical
  • More boring
  • More regulated
  • More embedded
That’s what maturation looks like.

Bottom line​

The AI bubble won’t burst like a balloon. It will leak.

  • Some valuations will collapse
  • Many startups will disappear
  • The hype will fade
But AI itself will likely become infrastructure—like electricity, cloud computing, or the internet—after the noise dies down.
 
KingOfPain said:
As a joke I asked Copilot: "When will the AI bubble burst?"

Short answer: it probably won’t “burst” all at once—but parts of it will almost certainly deflate.
Here’s a clear way to think about it.

1. What people mean by an “AI bubble”​

Most talk about an AI bubble isn’t about AI existing—models clearly work. It’s about:

  • Over-optimistic valuations of AI companies
  • Massive capital spending based on assumptions of rapid, universal adoption
  • Expectations that AI will quickly replace large parts of human labor
  • Revenue projections that assume users will pay much more than they currently do
That combination is classic bubble terrain.

2. Why a sudden crash is unlikely​

Unlike the dot‑com era, AI has several stabilizers:

✅ Real, immediate utility​

AI already improves productivity in:

  • Software development
  • Marketing and content creation
  • Customer support
  • Research and analysis
This means demand won’t vanish overnight.

✅ Deep-pocketed buyers​

Big tech firms (Microsoft, Google, Amazon, Meta) can absorb losses for years. Many AI investments are strategic, not speculative.

✅ Infrastructure lock-in​

Once AI is embedded into workflows (Office, Google Docs, CRMs, IDEs), companies don’t simply rip it out.

Because of this, there’s unlikely to be a dramatic, single “pop.”

3. Where the bubble​

Instead of a burst, expect multiple mini-corrections.

1. Startup shakeout (already happening)​

  • Many AI startups are thin wrappers around the same foundation models
  • Limited defensibility
  • High compute costs, weak differentiation
Expect:

  • Consolidation
  • Acquisitions at down-round prices
  • Quiet shutdowns

2. Valuation reality check (next 1–3 years)​

Investors will increasingly ask:

  • Can this company make money without subsidized compute?
  • Are customers retaining usage after novelty fades?
  • Is this saving money—or just sounding impressive?
Firms that answer poorly will reset hard.

3. “AI everywhere” fatigue​

Businesses are discovering:

  • AI works best with clean data (which most don’t have)
  • Human oversight remains expensive
  • Deployment is harder than demos suggest
This slows adoption curves and reduces hype-driven growth rates.

4. Historical parallels (and what they teach us)​

Dot-com bubble​

  • Web survived, thousands of companies didn’t
  • Amazon won, Pets.com lost

Smartphone boom​

  • Not everyone became Apple or Google
  • Margins collapsed for most hardware players
AI will likely follow the same pattern:


5. When does the “AI bubble” peak?​

If we define “burst” as expectations realigning with reality, a reasonable window is:

2026–2029

That’s when:

  • AI cost curves flatten
  • Revenue lags hype
  • Labor displacement proves slower than feared
  • Regulation starts adding friction
But note: this would be a cooling period, not a crash.

6. What probably will​

  • ❌ AI suddenly stops improving
  • ❌ Companies abandon AI wholesale
  • ❌ Models become useless or irrelevant
Instead, AI becomes:

  • Less magical
  • More boring
  • More regulated
  • More embedded
That’s what maturation looks like.

Bottom line​

The AI bubble won’t burst like a balloon. It will leak.

  • Some valuations will collapse
  • Many startups will disappear
  • The hype will fade
But AI itself will likely become infrastructure—like electricity, cloud computing, or the internet—after the noise dies down.
Click to expand...
I know this was a joke to see what ChatGPT would say, but to answer seriously:

Some of those exact reasons could also apply as to why the dotcom bubble wouldn’t burst all at once … and look how that turned out. The AI investments are much much bigger and it’s not clear the productivity gains are worth those investments:



Sadly I don’t have a WSJ subscription to read the whole article but yeah there’s a massive disconnect here between the corporate leaders and the people actually having to do the work. Sure it’s embedded everywhere, but often where it doesn’t need to be or at the very least doesn’t seem to be driving growth.

Again, not a hater of the technology per se, I think it can be useful in certain areas. But the idea that the AI economy is anything other than a massive bubble waiting to burst - one that could potentially eclipse the dotcom bubble - seems far fetched to this very non-financial expert. Heck given the size of investment, even a “minor” correction could be the size of the dotcom crash.
 
Last edited:
dada_dave said:
Some of those exact reasons could also apply as to why the dotcom bubble wouldn’t burst all at once … and look how that turned out. The AI investments are much much bigger and it’s not clear the productivity gains are worth those investments:
Click to expand...

I wanted to post a short comment, but ran out of time. Thanks for picking up the baton.

What irks me most about the Copilot repsonse is this part:
"Many AI investments are strategic, not speculative."

Just because one may call it "strategic", doesn't mean that the strategy has to be sound.
And pumping billions of dollars into an industry that still doesn't seem to have a clear path to profit sounds speculative to me.

As you correctly state, there is a big size difference between a dotcom company and what is going on here.
A dotcom company might have had a small server farm. AI companies have and plan warehouses full of servers that need the same power as a city and increase RAM costs for the whole market.
EDIT: Additionally, the other argument seems to be "it's too big to fail", which has never been a problem before...
 
Last edited:

Similar threads

Chew Toy McCoy
The death by AI thread
2
Replies
24
Views
1K
dada_dave
D
Eric
Jerome Powell says the AI hiring apocalypse is real: ‘Job creation is pretty close to zero’
Replies
7
Views
362
Hrafn
H
Chew Toy McCoy
The first AI actress
Replies
10
Views
480
dada_dave
D
Chew Toy McCoy
AI video sharing thread
Replies
2
Views
353
Chew Toy McCoy
Chew Toy McCoy
Chew Toy McCoy
The return of the random video thread
Replies
11
Views
875
Chew Toy McCoy
Chew Toy McCoy
Back
Top