AI will do a cost comparison between any Thai city or any city for that matter. Very handy if looking for a place to live.

3 hours ago, n8sail said:

I'll take you to task on this one. Normally I agree with your usually well-thought out responses here.

It depends what you're using it for and what model you're using. For average users of the internet, yes it is pretty stupid as it's mostly used for mundane photo edits and time-wasting videos. The in-built AI on all current modern web browsers suffices for this usage (for me as well)

I started using Claude Cowork 3 weeks ago. I am 17 years in the same technical manufacturing industry, 13 years here in Thailand at the same factory.

I can tell you Claude is transformational. I mean, REALLY transformational. My workload will be cut in half or less once I am fully up and running with it. I will not lose my job, it will just become easier and I will become even more knowledgeable as it sucks up all my thousands of pages of documentation and data and can parse it instantly when I have a problem to troubleshoot.

I have created an entire QC database in MS Azure to take the place of the spreadsheet-based system we have been using for many years now. When I graduated uni 20 years ago in IT Management, it would have taken me a few months of full-time labor to make this same relational object oriented database with multiple web-app UIs using MS Access. We're a small company and never before had the resources to attack a project like this. The modern Azure DB took me 3 days to get going.

Good friend of mine is software developer for large company here. 3 years ago they started developing their own LLM. 6 months ago we were hanging out. He said he does the work of 3+ experienced coders now, and in far less time.

That is NOT "Stupid"

Yes I would have to agree with you, and I should have qualified it by saying that the AI that most of us have free access to is fairly stupid, and as you stated I'm sure there are infinitely more advanced versions out there.

1 hour ago, save the frogs said:

AI cannot be fully trusted because in the health field, there are many contradictory studies. You will need to query it carefully ... all the best

Yes, need to pay attention to the source, that AI is getting the info from. Many times it's simply FB or YT. Wiki occasionally, better for med info than political.

It is not the technology itself. It is who is in control of it and to what end - PROFIT. I feel it is the same with the Internet and Smart phones. These are all great tools, tributes to the minds of humans, but to have this in the hands of companies with unprecedented wealth and political power gives me reason for pause.

We should all have a say about the future of a technology as powerful as AI, but set in the hands of a powerful few people.......

On 4/16/2026 at 9:11 AM, atpeace said:

It has had a huge impact on improving the quality of my life. I created 8 apps that I always wanted to track personal data from diet, health, money, and fitness. Doing the programming pre AI was always extremely tedious but now it is fun. I get to work through issues without all the infrastructure setup and nailing the expressions I need in whatever language. My mental clarity has noticeably increased. I'm running out of projects though :)

Also, I can have disagreements with AI and it doesn't shout at me like so many posters here. Also, it has no need to be in a herd which is possibly the best part. It is refreshing to be able to have a discussion online that resembles the civility of a real life face to face discussion with an individual.

Well as some of us know AI does not exits yet, it is a good complex machine learning system though.

I use ChatGPT for many hours each day when I design and sell my iPhone apps on the App Store but it still takes six months or a year to make a complex app which will be accepted on the Apple App Store. (By a single person that is). (30 years freelance software engineering experience).

Yes it can knock out noddy/simple apps on its own but nothing complex yet.

But it is good at creating simple outlines to apps and even good at checking the code we have written as would happen in a code review by other software engineers.

We can use it in an iterative way to build up code as long as we know in detail what the code is doing and we correct it or get ChatGPT to correct it when we spot errors.

But we still have to tell it the detailed requirements and build up the app as we go.

You can test that by asking it to build you a fully implemented banking app, with the UI in place, the encryption in place, the front end, the back end banking system in the cloud etc and see what it says, it will give you a simple outline only and can not do the rest.

Even ask it the question "are you real AI by the real definition of AI" and it will tell you it is not.

It makes mistakes all the time but is open to suggestion and corrections.

It will not replace software engineers in the same way Auto-cad which was invented decades ago has not replaced architects.

Most people do not know what software engineers are and think a programmed system is created by people called coders as seen in films, but coding is only a very small part of software engineering, films where a coder creates a who system by typing away at the keyboard and magically creates a program makes my laugh.

But as I said it is an excellent tool.

It can be used I suppose by doctors, accountants etc top check their work as a cross reference but not as a stand alone machine, and a machine (or a set of machines) is what it is.

We are probably a hundred years away from real AI, but the investors, people selling so called AI have hyped up the market, it will burst in the not to distant future and a lot of people will have lost out in their investments just like its the .com bubble burst.

But overall it is a great tool as long as people know how to use it and always cross reference the results.

And it is not doom and gloom as the people who do not even know what it is or can do say, they are afraid of anything new or something they do not understand.

On 4/16/2026 at 9:58 AM, Rockyroad said:

I ask AI about workout routines and diet. It saves time.

You will have to stop writing positive comments on this blog as you already have five thumbs down by the miseries, I like dropping by on this blog now and then for a break from 'work' as the miseries make me laugh, I have had hundreds of thumbs down from them for writing positive things, it gives them some sort of meaning in life I suppose. 😀

22 minutes ago, JamesPhuket10 said:

But as I said it is an excellent tool.

I'm not doing any coding but I though there were specialized apps for coders, not ChatGPT ?

ie GitHub Copilot?

Anyway, no need to respond. I'm not interested in going down the technical path of software coding as I am not a coder.

https://github.com/features/copilot

On 4/16/2026 at 3:29 PM, save the frogs said:

some people will not use AI much or at all, people living in remote villages ..

Haha, I live in a remote village and I have a techhub here, supplied by high-speed fibre optic internet. And I use AI extensively...

3 hours ago, atpeace said:

It does but the power of AI is mind boggling. I can't wrap my head around the potential. It takes some effort to learn how to use it and everybody will have unique needs. If I had to choose between my beautiful caring wife and AI, I would still choose her but the gap is closing.

This morning I had a two our discussion with AI about my blood work and fluid retention after running and sitting for hours. I had my theories and AI had its theories. In the end, I was able to design a plan that should definitely improve upon some blood numbers and fluid retention. I'm a health nut and know more about health than most but compared to AI, I'm a fool. If you ask the right questions and provide the correct context, it can do amazing things.

3 hours ago, atpeace said:

It does but the power of AI is mind boggling. I can't wrap my head around the potential. It takes some effort to learn how to use it and everybody will have unique needs. If I had to choose between my beautiful caring wife and AI, I would still choose her but the gap is closing.

This morning I had a two our discussion with AI about my blood work and fluid retention after running and sitting for hours. I had my theories and AI had its theories. In the end, I was able to design a plan that should definitely improve upon some blood numbers and fluid retention. I'm a health nut and know more about health than most but compared to AI, I'm a fool. If you ask the right questions and provide the correct context, it can do amazing things.

Did you not try the internet previously ?

23 hours ago, b17 said:

I was having a back and forth with Chat GPT about the upcoming NHL series between Montreal and Tampa Bay, and was astonished when it began spouting names that haven't been on the Montreal squad for over a year. I corrected it, it apologized, and then added two further names of former players. So, one might say a developing technology, yet imperfect to a degree that is obvious to people aware of a subject.

I've tried ChatGPT and Gemini extensively. I have now settled on a paid subscription to Gemini because ChatGPT did similar things to me.

4 hours ago, captain_shane said:

This is the most glaringly ignorant statement in the entire text, demonstrating a complete lack of understanding of basic thermodynamics. The vacuum of space is an insulator, not a coolant. Because there's no matter in a vacuum to conduct or convect heat away, spacecraft must rely entirely on radiative cooling, which is incredibly inefficient. A datacenter generating the massive thermal output required for "serious AI" would melt itself into slag in orbit without impossibly massive, highly vulnerable radiator panels. You also ignore cosmic radiation (which causes bit-flips and destroys unshielded microprocessors), latency constraints (the speed of light creates hard limits for data transmission to Earth), and the exorbitant cost of lifting millions of tons of heavy compute infrastructure into orbit. 10-15 years? This is a fantasy completely divorced from aerospace engineering realities.

I'm not even going to go into your complete lack of understanding what software engineers do. Obsolete? Lol, you have no idea what you're talking about.

Orbital, SpaceX and Blue Origin are all working on space data centres. Ten years is a long time to these companies.

BTW, you're right that cooling is challenging because convection doesn't work in space but with shade temperatures of -170 deg. C and liquid cooling, they'll get there.

1 minute ago, IsaanT said:

Orbital, SpaceX and Blue Origin are all working on space data centres. Ten years is a long time to these companies.

BTW, you're right that cooling is challenging because convection doesn't work in space but with shade temperatures of -170 deg. C and liquid cooling, they'll get there.

Not gonna happen. You can't "engineer around" the Stefan-Boltzmann law. To cool a multi-gigawatt AI datacenter in the vacuum of space, it would require radiator panels with surface areas the size of small cities. This also completely ignores maintenance. GPUs and servers fail constantly. In an earth-based datacenter, technicians swap out dead hardware daily. Who's swapping out the thousands of degraded NVIDIA chips in your orbital datacenter? Are you flying a Starship with human technicians to LEO every Tuesday to replace burnt-out RAM? The operational logistics make this economically unjustifiable.

7 minutes ago, captain_shane said:

Not gonna happen. You can't "engineer around" the Stefan-Boltzmann law. To cool a multi-gigawatt AI datacenter in the vacuum of space, it would require radiator panels with surface areas the size of small cities. This also completely ignores maintenance. GPUs and servers fail constantly. In an earth-based datacenter, technicians swap out dead hardware daily. Who's swapping out the thousands of degraded NVIDIA chips in your orbital datacenter? Are you flying a Starship with human technicians to LEO every Tuesday to replace burnt-out RAM? The operational logistics make this economically unjustifiable.

Given their track records, would you bet against Musk- and Bezos-owned companies...? One is the richest man on the planet, the other is currently the fourth richest. Just saying...

Just now, IsaanT said:

Given their track records, would you bet against Musk- and Bezos-owned companies...?

Yes lol. Musk gave up on mars, because that was also a stupid idea. Same as his stupid excavation company and many others. Bezos' ai is so <deleted>ty it took down aws last week. ZERO chance we have space data centers in 10 years. These same idiots were claiming we'd have agi in 2025 and Musk said we'd have full on recursive self-learning superintelligence by now lol.

1 hour ago, IsaanT said:

Haha, I live in a remote village and I have a techhub here, supplied by high-speed fibre optic internet. And I use AI extensively...

I am in Phuket, we can be anywhere if we have wifi etc, then all we need is a brain to know how to use AI effectively 👍

Use Grok to explain Paywall blocked articles and general info.

1 hour ago, save the frogs said:

I'm not doing any coding but I though there were specialized apps for coders, not ChatGPT ?

ie GitHub Copilot?

Anyway, no need to respond. I'm not interested in going down the technical path of software coding as I am not a coder.

https://github.com/features/copilot

There is no such thing as a coder, they only exist on daft films. 😃

Superman etc also do not exist in real life.

ChatGPT is excellent for software engineering, I use it every day.

It is also great for non engineering topics as well, I had a chat with it about how so many bar girls from Isaan managed to snare daft farangs into buying a house, car and pay for her parents upkeep etc and it was aware of all of their tricks and the psychology they use over the old guys to get to their money etc.

It has been obvious to me over the years how it works, it seemed to know how it does work which agreed with my knowledge of how they operate.

A bit off topic but it shows as an example of how it can be used in many different ways.

1 hour ago, JamesPhuket10 said:

Well as some of us know AI does not exits yet, it is a good complex machine learning system though.

While LLMs are not true AI, they are very useful. I might even consider paying a monthly fee for a model.

23 minutes ago, JamesPhuket10 said:

about how so many bar girls from Isaan managed to snare daft farangs into buying a house, car and pay for her parents upkeep etc

Ha, you're obsessed with that aren't you?

You must miss Cameroni.

The girls don't need psychological tricks. Men want to get laid and don't even mind parting with their money to get laid.

Men have an innate provider instinct, and so it works out.

On 4/16/2026 at 3:29 PM, save the frogs said:

some people will not use AI much or at all, people living in remote villages ..

some people will go the other extreme and become cyborgs ...

danger of your own mental faculties declining if you outsource all your decisions to AI and let AI write all your blog posts on AN.

I would never do that.

What would be the point.

Anyway, it must be so obvious to anyone that I have never used, and will never use, an AI to write any Gamma comment.

You could not pay enough to stoop so low.

I don't like comments that use AI or are totally AI generated.

Responding to such comments would be like having sex with one of those blow-up dolls.

Where is the humanity in that?

3 hours ago, JamesPhuket10 said:

You will have to stop writing positive comments on this blog as you already have five thumbs down by the miseries, I like dropping by on this blog now and then for a break from 'work' as the miseries make me laugh, I have had hundreds of thumbs down from them for writing positive things, it gives them some sort of meaning in life I suppose. 😀

They have not seen a gym in 10 years. Probably can't even walk to the shops and back.

I asked Gemini to summarise this thread....

Here’s a clear, no-nonsense summary of that thread:

---

🧠 Overall Theme

The discussion splits into three broad camps:

1. Practical users (pro-AI)

2. Skeptics (limitations & risks)

3. Alarmists (AI is dangerous/evil)

---

👍 How people say AI is improving their lives

Several users report real, practical benefits:

• Productivity & automation

  • Building apps, coding, solving technical problems much faster.

  • Automating everyday tasks (taxes, finance, cooking, dog training).

• Information & learning

  • Replacing Google searches with AI answers.

  • Used for research, fitness advice, diet planning.

• Health discussions

  • Some use AI to explore medical issues in depth (though not always trusting doctors alone).

• Cost & time savings

  • No need for personal trainers, consultants, or some professional services.

• Better interaction experience

  • AI is seen as calm, non-judgmental, and not “tribal” like online forums.

---

😐 Common criticisms & limitations

Even regular users acknowledge flaws:

• Inaccuracies

  • AI gives outdated or incorrect information (e.g., wrong sports rosters).

  • Requires users to double-check facts.

• Not truly “beyond the hype” yet

  • Still relies on the same messy internet data—good, bad, and biased.

• Content quality decline

  • AI blamed for flooding platforms like YouTube with low-quality content.

• Censorship & restrictions

  • Some frustration that AI refuses certain requests or deletes conversations.

---

⚠️ Concerns about risks

A number of posters raise cautionary points:

• Privacy fears

  • Worry that AI tools access or misuse personal data.

  • Advice: don’t share anything sensitive.

• Over-reliance

  • Risk of people losing critical thinking or “mental faculties.”

• Security & control

  • Suspicion about big tech and data ownership.

---

😨 Strong negative / extreme views

A smaller but loud group sees AI as:

• A major threat to humanity

• Comparable to a plague or scourge

• Something that could radically disrupt civilization within decades

Others push back, comparing this fear to historical resistance to new technology (e.g., Industrial Revolution).

---

🔥 Tone of the discussion

• Mix of thoughtful insights and typical forum chaos.

• Some useful real-world examples, but also:

  • Personal attacks

  • Conspiracy-style thinking

  • Off-topic rants (including about AI leaders)

---

🧩 Bottom line

• AI is already useful in everyday life, especially for saving time and expanding capabilities.

• It’s not fully reliable, and users know it.

• Public opinion is sharply divided—from enthusiasm to deep suspicion to outright fear.

---

Here’s the sharper takeaway—the signal underneath all the noise:

---

🧠 What this thread actually reveals about how people see AI right now

1. AI is already normal—but not fully trusted

People are using AI for everyday things (taxes, fitness, coding, research), which means it’s crossed the line from novelty to utility.

But—almost every user adds a caveat:

“It’s useful… but I double-check it.”

That’s the current mindset:
👉 “Helpful assistant, not an authority.”

---

2. The real value is time, not intelligence

Nobody is amazed by AI being “smart.”

What they actually value:

• Saving time

• Reducing effort

• Replacing tedious tasks

👉 AI isn’t seen as genius—it’s seen as a labour-saving device.

---

3. Power users are pulling ahead quietly

One standout pattern:

• Casual users → asking questions, basic tasks

• Advanced users → building apps, automating life, thinking better

That gap is widening.

👉 AI isn’t just a tool—it’s becoming a multiplier of personal capability, and not everyone is keeping up.

---

4. Mistrust is everywhere—but unfocused

People are uneasy, but not in a precise way. Concerns are scattered:

• “It might steal my data”

• “It’s censored”

• “It might destroy society”

• “Tech companies are dodgy”

These aren’t coordinated fears—they’re ambient suspicion.

👉 The vibe is: “Something big is happening… not sure I like it.”

---

5. People project human traits onto AI

This comes up repeatedly:

• “It doesn’t argue”

• “It’s calm”

• “It adapts to me”

• “It apologizes”

Users are treating AI less like software and more like a personality interface.

👉 That’s a deeper shift than most realise.

---

6. Accuracy problems are obvious—but tolerated

Everyone notices mistakes.

But no one stops using it.

👉 This is crucial:
People have already decided:

“Imperfect but useful” beats “accurate but slow.”

---

7. There’s a quiet divide forming

You can see three emerging groups:

• Adapters → integrating AI into daily life

• Skeptics → cautious, selective use

• Rejecters → distrust it completely

👉 This isn’t just opinion—it’s the early stages of a behavior gap that will matter later.

---

🧩 The real bottom line

This thread shows we’re in a very specific phase:

AI is no longer impressive—but it’s not fully trusted either.

People aren’t asking “What is this?” anymore.

They’re asking:

• “How much should I rely on it?”

• “What’s the catch?”

• “Am I gaining an edge—or being fooled?”

---

⚡ One-line distillation

👉 AI has quietly become useful enough that people keep using it—even while not quite believing in it.

---

4 hours ago, JamesPhuket10 said:

You will have to stop writing positive comments on this blog as you already have five thumbs down by the miseries, I like dropping by on this blog now and then for a break from 'work' as the miseries make me laugh, I have had hundreds of thumbs down from them for writing positive things, it gives them some sort of meaning in life I suppose. 😀

It is incredible insane. They get a rush out of following him around and downvoting every post. What evil jerks...

3 hours ago, JamesPhuket10 said:

It is also great for non engineering topics as well, I had a chat with it about how so many bar girls from Isaan managed to snare daft farangs into buying a house, car and pay for her parents upkeep etc and it was aware of all of their tricks and the psychology they use over the old guys to get to their money etc.

Don't tell Georgegeorgea, he'll be devastated and hang up his hi vis shirt, pens and notebooks for good.

36 minutes ago, Dexxter said:

I asked Gemini to summarise this thread....

Here’s a clear, no-nonsense summary of that thread:

---

🧠 Overall Theme

The discussion splits into three broad camps:

1. Practical users (pro-AI)

2. Skeptics (limitations & risks)

3. Alarmists (AI is dangerous/evil)

---

👍 How people say AI is improving their lives

Several users report real, practical benefits:

• Productivity & automation

  • Building apps, coding, solving technical problems much faster.

  • Automating everyday tasks (taxes, finance, cooking, dog training).

• Information & learning

  • Replacing Google searches with AI answers.

  • Used for research, fitness advice, diet planning.

• Health discussions

  • Some use AI to explore medical issues in depth (though not always trusting doctors alone).

• Cost & time savings

  • No need for personal trainers, consultants, or some professional services.

• Better interaction experience

  • AI is seen as calm, non-judgmental, and not “tribal” like online forums.

---

😐 Common criticisms & limitations

Even regular users acknowledge flaws:

• Inaccuracies

  • AI gives outdated or incorrect information (e.g., wrong sports rosters).

  • Requires users to double-check facts.

• Not truly “beyond the hype” yet

  • Still relies on the same messy internet data—good, bad, and biased.

• Content quality decline

  • AI blamed for flooding platforms like YouTube with low-quality content.

• Censorship & restrictions

  • Some frustration that AI refuses certain requests or deletes conversations.

---

⚠️ Concerns about risks

A number of posters raise cautionary points:

• Privacy fears

  • Worry that AI tools access or misuse personal data.

  • Advice: don’t share anything sensitive.

• Over-reliance

  • Risk of people losing critical thinking or “mental faculties.”

• Security & control

  • Suspicion about big tech and data ownership.

---

😨 Strong negative / extreme views

A smaller but loud group sees AI as:

• A major threat to humanity

• Comparable to a plague or scourge

• Something that could radically disrupt civilization within decades

Others push back, comparing this fear to historical resistance to new technology (e.g., Industrial Revolution).

---

🔥 Tone of the discussion

• Mix of thoughtful insights and typical forum chaos.

• Some useful real-world examples, but also:

  • Personal attacks

  • Conspiracy-style thinking

  • Off-topic rants (including about AI leaders)

---

🧩 Bottom line

• AI is already useful in everyday life, especially for saving time and expanding capabilities.

• It’s not fully reliable, and users know it.

• Public opinion is sharply divided—from enthusiasm to deep suspicion to outright fear.

---

Here’s the sharper takeaway—the signal underneath all the noise:

---

🧠 What this thread actually reveals about how people see AI right now

1. AI is already normal—but not fully trusted

People are using AI for everyday things (taxes, fitness, coding, research), which means it’s crossed the line from novelty to utility.

But—almost every user adds a caveat:

That’s the current mindset:
👉 “Helpful assistant, not an authority.”

---

2. The real value is time, not intelligence

Nobody is amazed by AI being “smart.”

What they actually value:

• Saving time

• Reducing effort

• Replacing tedious tasks

👉 AI isn’t seen as genius—it’s seen as a labour-saving device.

---

3. Power users are pulling ahead quietly

One standout pattern:

• Casual users → asking questions, basic tasks

• Advanced users → building apps, automating life, thinking better

That gap is widening.

👉 AI isn’t just a tool—it’s becoming a multiplier of personal capability, and not everyone is keeping up.

---

4. Mistrust is everywhere—but unfocused

People are uneasy, but not in a precise way. Concerns are scattered:

• “It might steal my data”

• “It’s censored”

• “It might destroy society”

• “Tech companies are dodgy”

These aren’t coordinated fears—they’re ambient suspicion.

👉 The vibe is: “Something big is happening… not sure I like it.”

---

5. People project human traits onto AI

This comes up repeatedly:

• “It doesn’t argue”

• “It’s calm”

• “It adapts to me”

• “It apologizes”

Users are treating AI less like software and more like a personality interface.

👉 That’s a deeper shift than most realise.

---

6. Accuracy problems are obvious—but tolerated

Everyone notices mistakes.

But no one stops using it.

👉 This is crucial:
People have already decided:

---

7. There’s a quiet divide forming

You can see three emerging groups:

• Adapters → integrating AI into daily life

• Skeptics → cautious, selective use

• Rejecters → distrust it completely

👉 This isn’t just opinion—it’s the early stages of a behavior gap that will matter later.

---

🧩 The real bottom line

This thread shows we’re in a very specific phase:

People aren’t asking “What is this?” anymore.

They’re asking:

• “How much should I rely on it?”

• “What’s the catch?”

• “Am I gaining an edge—or being fooled?”

---

⚡ One-line distillation

👉 AI has quietly become useful enough that people keep using it—even while not quite believing in it.

---

More like War and Peace than a summary.

10 hours ago, spidermike007 said:

AI is so stupid it can't even handle the most simple task like dictation, without making a half dozen mistakes in one paragraph. I think for the most part we're only getting access to the $6 version at this point.

I think you might be confusing google with AI.

I just used Chat GTP last week to review and revise a contract I wrote prior to sending it to my lawyer.

When my lawyer received it he said what I sent him covered all the bases and was good to go.

My brother just implemented an AI agent (bot) for our insurance agency 3 months ago.

It's amazing, This bot sets up appointments for our agents. We have a CRM that the Bot uses to send text messages to leads. We set how many leads per day and off it goes.

The bot reaches out to potential clients, answers their questions, and will schedule the appointment with the client. It will then add that appointment to the agents calendar.

So instead of having 3 girls on the phone calling all day we have an AI agent / Bot or whatever you want to call it setting up the appointments.

Our appointment rate went up from 68% using the girls to 74% with the AI bot . The girls that were in the office are now out selling.

No hustling and trying to convince someone to meet with you. Just show up for the scheduled appointment and close it. Our agents closer rate is still the same 98%

13 hours ago, spidermike007 said:

AI is so stupid it can't even handle the most simple task like dictation, without making a half dozen mistakes in one paragraph. I think for the most part we're only getting access to the $6 version at this point.

Lot's of fear mongering but yes in 10 years the world will be a super different place.

Classism will be its most profound effect on society.

AI is in its infancy now and you ain't seen nothing yet.

14 hours ago, captain_shane said:

Not gonna happen. You can't "engineer around" the Stefan-Boltzmann law. To cool a multi-gigawatt AI datacenter in the vacuum of space, it would require radiator panels with surface areas the size of small cities. This also completely ignores maintenance. GPUs and servers fail constantly. In an earth-based datacenter, technicians swap out dead hardware daily. Who's swapping out the thousands of degraded NVIDIA chips in your orbital datacenter? Are you flying a Starship with human technicians to LEO every Tuesday to replace burnt-out RAM? The operational logistics make this economically unjustifiable.

The plan isn't for a single giant datacenter. Musk wants to have millions of smaller satellites and distribute the load. He already has thousands of satellites up there and has learned how to collect solar energy.

Regarding maintenance - when a part fails, the satellite is decommissioned. It falls back to Earth and burns up in the atmosphere, and Musk launches a replacement.

Right now it's too expensive to launch that many, but they are trying to address all that.

On 4/16/2026 at 2:49 AM, spidermike007 said:

AI is the most dangerous thing to happen to humanity in the last 50 years, and 10 to 20 years from now we are not going to recognize civilization as a result of the extraordinarily dangerous implications that AI represents.

AI is a plague

AI is a scourge

AI is absolutely horrific

The rise of AI poses a serious threat to our world and could potentially lead to the downfall of human civilization. I’m scared.

What is stupid is your belief that you wont lose your job one day because of AI.A =n australian based tech company recently laid off 1400 staff citing AI.

18 hours ago, atpeace said:

It does but the power of AI is mind boggling. I can't wrap my head around the potential. It takes some effort to learn how to use it and everybody will have unique needs. If I had to choose between my beautiful caring wife and AI, I would still choose her but the gap is closing.

This morning I had a two our discussion with AI about my blood work and fluid retention after running and sitting for hours. I had my theories and AI had its theories. In the end, I was able to design a plan that should definitely improve upon some blood numbers and fluid retention. I'm a health nut and know more about health than most but compared to AI, I'm a fool. If you ask the right questions and provide the correct context, it can do amazing things.

I'll take you to task on this one. Normally I agree with your usually well-thought out responses here.

It depends what you're using it for and what model you're using. For average users of the internet, yes it is pretty stupid as it's mostly used for mundane photo edits and time-wasting videos. The in-built AI on all current modern web browsers suffices for this usage (for me as well)

I started using Claude Cowork 3 weeks ago. I am 17 years in the same technical manufacturing industry, 13 years here in Thailand at the same factory.

I can tell you Claude is transformational. I mean, REALLY transformational. My workload will be cut in half or less once I am fully up and running with it. I will not lose my job, it will just become easier and I will become even more knowledgeable as it sucks up all my thousands of pages of documentation and data and can parse it instantly when I have a problem to troubleshoot.

I have created an entire QC database in MS Azure to take the place of the spreadsheet-based system we have been using for many years now. When I graduated uni 20 years ago in IT Management, it would have taken me a few months of full-time labor to make this same relational object oriented database with multiple web-app UIs using MS Access. We're a small company and never before had the resources to attack a project like this. The modern Azure DB took me 3 days to get going.

19 hours ago, n8sail said:

Good friend of mine is software developer for large company here. 3 years ago they started developing their own LLM. 6 months ago we were hanging out. He said he does the work of 3+ experienced coders now, and in far less time.

That is NOT "Stupid"

18 hours ago, atpeace said:

It does but the power of AI is mind boggling. I can't wrap my head around the potential. It takes some effort to learn how to use it and everybody will have unique needs. If I had to choose between my beautiful caring wife and AI, I would still choose her but the gap is closing.

This morning I had a two our discussion with AI about my blood work and fluid retention after running and sitting for hours. I had my theories and AI had its theories. In the end, I was able to design a plan that should definitely improve upon some blood numbers and fluid retention. I'm a health nut and know more about health than most but compared to AI, I'm a fool. If you ask the right questions and provide the correct context, it can do amazing things.

18 hours ago, atpeace said:

It does but the power of AI is mind boggling. I can't wrap my head around the potential. It takes some effort to learn how to use it and everybody will have unique needs. If I had to choose between my beautiful caring wife and AI, I would still choose her but the gap is closing.

This morning I had a two our discussion with AI about my blood work and fluid retention after running and sitting for hours. I had my theories and AI had its theories. In the end, I was able to design a plan that should definitely improve upon some blood numbers and fluid retention. I'm a health nut and know more about health than most but compared to AI, I'm a fool. If you ask the right questions and provide the correct context, it can do amazing things.

You've got to bear in mind that AI is only a computer programme and it only knoes what is fed into it and if bullsh@t is fed into it then that is what you get out of it. I would advise you to have a discussion with a qualified doctor instead of a computer as your condition sounds serious instead of a computer.