I’m 91 and you have saved me from the despair of a future for the coming generations. AI with energy from SMRs and education that the young will love to engage with give me confidence that our future may be bright after all.
I have used Waymo when I've visited Austin and I love it! No potentially driver, no music, no need to tip, and Waymos are so damn cute! I wish they were everywhere. I've read a little about Alpha schools and know a few people to who are involved with them. I think they are amazing from everything I've learned about them. One small observation: you can never entirely get rid of drills and memorization. Kids need to know multiplication tables, rules of grammar, etc. I still use this information in my seventies. There is information that can save time if it's available in your data banks.
Your optimism is misplaced. I'm all for innovation and efficiency; those things were never the problem. The root problem is selfish human nature. Perhaps AI can be made "three laws safe," but someone will always find a workaround for that. As I am fond of saying, You can't fix a problem caused by selfish human nature with the same selfish nature that caused the problem in the first place. I give the solution to this in Discourse On Agape, read free here: https://discourseonagape.org/booklets/ Relatively few people will avail themselves of the solution, however.
So they say. But 'rational self interest' has no morals, by definition, and can, and has been, used to justify the most horrific acts, which to the perpetrator seems rational, at the time.
When you say, "So they say," who is "they?" Rational self interest is a code of morality. If someone's rights are violated, then the action wasn't rational. What is an example of a "horrific" action that you think was driven by rational self-interest?
The 'Randians,' is one group that says that. Many atheists will also argue they have 'rational self interest,' and they proved it in China and Russia. Gaza comes to mind as well. To be fair, jews did produce the documentary Tantura, so they're not all bad. https://archive.org/details/tantura_2022
Others rights were violated, but, of course, 'it was for the greater good.' All people naturally believe in 'rational self interest,' which includes the right of self defense, but is the world in any better shape than it was a few thousand years ago? Man has progressed in technology, but instead of using it to make the world a better place, he uses it to kill his fellow man - by the hundreds of millions. The last century was the wealthiest, and bloodiest, in known history. We are headed for WWIII, since WWII didn't really solve anything. We can go to sleep at night and forget our troubles for a few hours. God can't do that; He has to watch His children destroy themselves and others, because they have 'rational self interest,' rather than receiving His spirit of agape. Say a prayer for Him.
I can't help but think that despite the obvious benefits, things like robotaxi are just one more step towards a future where the individual gives up more and more autonomy. I'm not making an argument, just expressing an uneasy feeling that we are giving up our agency in the name of convenience.
He compared seatbelts to Waymo. Agreed, no sane person would deny that seatbelts are good, until they become mandatory. When the local cop can crack your skull for not using one, we have a problem. Waymo is good, but when/if it becomes mandatory to use one, it's time to do some housekeeping. Every totalitarian idea can be rationalized in order to keep us "safe," see the Covid hysteria, the ACA, and the Patriot Act for recent examples.
First chart should include number of nukes/SMRs in Rickover's Navy; designed, built & installed AFTER NRC. Just read Breakneck by David Wang & Abundance by Derek Thompson and Ezra Klein that gets into why China makes more. One of the answers is some people make things happen. This is George S. Patton -- Lead, Follow or get out of the way. Which you verify 2 charts later that shows the significant cost reduction to launch with SpaceX & Starship.
The tech people in Austin supporting fractal computing contend that the fractal approach would decentralize the computing process and would reduce the need for centralized AI data centers. Can you comment?
The US NRC issued 14 Combined Construction and Operating Licenses (COL) between 1982 and 2025 under 10 Code of Federal Regulations Part 52 with one application currently under review (asof 1st Qtr 2026). Two of those COLs were constructed, Vogtle, six of those COL's are still active for construction.
We’ve reached the event horizon of AI noise. Our screens are becoming cluttered with 'perfect' content that has zero soul.
I believe the next great era isn't about more ,AI it’s about Purification. Using the machine to strip away the junk so we can return to high-signal, human-centric engineering and media.
I’m building SoloArchitect to be that filter. Leverage the AI, but protect the human. If you're ready for a cleaner, higher-signal version of the web, join the build.
The Alpha school situation is fraught with problems. Yes, AI has the potential to greatly improve education, but the implementation by the Alpha school model is a failure. Many parents are pulling out their children because of the intrusive monitoring of students even at home and behavioral problems created by the approach. A systemic or production line model of education will not work equally well for all students. All alpha does is apply a systemic process using AI. More study and better implementation are required. As a retired research director who has overseen both successes and failures, I consider myself a rational optimist but grounded in pragmatism, which requires one to look at all aspects of an issue not just be a cheerleader. This would be my primary criticism of your reporting. Please provide a more balanced view of your topics, such an approach would be more useful.
When Jimmy Kimmel showed Trump saying, "It's all computers" the audience laughed because it was a clueless old man who never drives himself.
It is odd to talk about Waymo and Tesla robotaxis as competiting. Waymo has been driverless since 2020. Zoox started in Vegas in October. Tesla announced it in Austin, but nobody has see one yet. Except the ones that have safety cars following.
SpaceX is amazing but xAI was merged with it because Grok is truly the worst of the LLMs. It doesn't matter how fast you build a data center if the LLM sucks.
I love the idea of the Alpha school. But do wonder how much of the success is self selective and probably good teachers. My kids use ixl and it is just a glorified quiz. Khan Academy is good but literally recordings of Khan lecturing.... not AI.
In cae you are wondering, Trump is the reason drones are falling on Abu Dabi.
Not surprised by the problem of the SF Bay Area data centers getting grid power. Place overall is notoriously left-leaning, overly concerned about the environment. Believe the small nuclear power plants will have a tremendous market. We don't have Waymos here, doubt I would ever use one unless I was no longer able to drive. Far as AI goes, not only do I avoid AI but don't use any 'smart' appliances unless they need not be hooked up to the Internet. Simply do not trust AI and 'smart' things in general. Have the background, was trying out VR glasses nearly forty years ago, was acquisition manager for two, later four large-scale computer-based training systems. AI is just one more tool, will never be a panacea, again, saw and heard all this stuff nearly forty years ago. Don't forget we need to put everything in perspective. Putting data centers in space is nice, but encryption of data streams going to and from the data centers will be paramount, and any encryption can be broken, just a matter of resources.
You recommend playing around with AI learning tools, do you have some recommendation with links that we can use to get started? I really enjoy your newsletter, I learn something new from every issue. Keep them coming.
It's healthy to feel optimistic about the future, given all the negativism in our world today.
Gives one hope for children and grandchildren.
I’m 91 and you have saved me from the despair of a future for the coming generations. AI with energy from SMRs and education that the young will love to engage with give me confidence that our future may be bright after all.
I have used Waymo when I've visited Austin and I love it! No potentially driver, no music, no need to tip, and Waymos are so damn cute! I wish they were everywhere. I've read a little about Alpha schools and know a few people to who are involved with them. I think they are amazing from everything I've learned about them. One small observation: you can never entirely get rid of drills and memorization. Kids need to know multiplication tables, rules of grammar, etc. I still use this information in my seventies. There is information that can save time if it's available in your data banks.
Your optimism is misplaced. I'm all for innovation and efficiency; those things were never the problem. The root problem is selfish human nature. Perhaps AI can be made "three laws safe," but someone will always find a workaround for that. As I am fond of saying, You can't fix a problem caused by selfish human nature with the same selfish nature that caused the problem in the first place. I give the solution to this in Discourse On Agape, read free here: https://discourseonagape.org/booklets/ Relatively few people will avail themselves of the solution, however.
Rational self-interest is good. It leads to human flourishing.
So they say. But 'rational self interest' has no morals, by definition, and can, and has been, used to justify the most horrific acts, which to the perpetrator seems rational, at the time.
When you say, "So they say," who is "they?" Rational self interest is a code of morality. If someone's rights are violated, then the action wasn't rational. What is an example of a "horrific" action that you think was driven by rational self-interest?
The 'Randians,' is one group that says that. Many atheists will also argue they have 'rational self interest,' and they proved it in China and Russia. Gaza comes to mind as well. To be fair, jews did produce the documentary Tantura, so they're not all bad. https://archive.org/details/tantura_2022
Others rights were violated, but, of course, 'it was for the greater good.' All people naturally believe in 'rational self interest,' which includes the right of self defense, but is the world in any better shape than it was a few thousand years ago? Man has progressed in technology, but instead of using it to make the world a better place, he uses it to kill his fellow man - by the hundreds of millions. The last century was the wealthiest, and bloodiest, in known history. We are headed for WWIII, since WWII didn't really solve anything. We can go to sleep at night and forget our troubles for a few hours. God can't do that; He has to watch His children destroy themselves and others, because they have 'rational self interest,' rather than receiving His spirit of agape. Say a prayer for Him.
The link to your guide to using AI doesn’t work.
1 Million AI satellites?!! Kessler Syndrome????
Exactly.
Stephen,
I can't help but think that despite the obvious benefits, things like robotaxi are just one more step towards a future where the individual gives up more and more autonomy. I'm not making an argument, just expressing an uneasy feeling that we are giving up our agency in the name of convenience.
He compared seatbelts to Waymo. Agreed, no sane person would deny that seatbelts are good, until they become mandatory. When the local cop can crack your skull for not using one, we have a problem. Waymo is good, but when/if it becomes mandatory to use one, it's time to do some housekeeping. Every totalitarian idea can be rationalized in order to keep us "safe," see the Covid hysteria, the ACA, and the Patriot Act for recent examples.
Wound me up more than usual!!!
First chart should include number of nukes/SMRs in Rickover's Navy; designed, built & installed AFTER NRC. Just read Breakneck by David Wang & Abundance by Derek Thompson and Ezra Klein that gets into why China makes more. One of the answers is some people make things happen. This is George S. Patton -- Lead, Follow or get out of the way. Which you verify 2 charts later that shows the significant cost reduction to launch with SpaceX & Starship.
The tech people in Austin supporting fractal computing contend that the fractal approach would decentralize the computing process and would reduce the need for centralized AI data centers. Can you comment?
The US NRC issued 14 Combined Construction and Operating Licenses (COL) between 1982 and 2025 under 10 Code of Federal Regulations Part 52 with one application currently under review (asof 1st Qtr 2026). Two of those COLs were constructed, Vogtle, six of those COL's are still active for construction.
We’ve reached the event horizon of AI noise. Our screens are becoming cluttered with 'perfect' content that has zero soul.
I believe the next great era isn't about more ,AI it’s about Purification. Using the machine to strip away the junk so we can return to high-signal, human-centric engineering and media.
I’m building SoloArchitect to be that filter. Leverage the AI, but protect the human. If you're ready for a cleaner, higher-signal version of the web, join the build.
I’m glad to see others are actually excited for the development of AI in ways to better the world in ways other than just increasing productivity.
The Alpha school situation is fraught with problems. Yes, AI has the potential to greatly improve education, but the implementation by the Alpha school model is a failure. Many parents are pulling out their children because of the intrusive monitoring of students even at home and behavioral problems created by the approach. A systemic or production line model of education will not work equally well for all students. All alpha does is apply a systemic process using AI. More study and better implementation are required. As a retired research director who has overseen both successes and failures, I consider myself a rational optimist but grounded in pragmatism, which requires one to look at all aspects of an issue not just be a cheerleader. This would be my primary criticism of your reporting. Please provide a more balanced view of your topics, such an approach would be more useful.
When Jimmy Kimmel showed Trump saying, "It's all computers" the audience laughed because it was a clueless old man who never drives himself.
It is odd to talk about Waymo and Tesla robotaxis as competiting. Waymo has been driverless since 2020. Zoox started in Vegas in October. Tesla announced it in Austin, but nobody has see one yet. Except the ones that have safety cars following.
SpaceX is amazing but xAI was merged with it because Grok is truly the worst of the LLMs. It doesn't matter how fast you build a data center if the LLM sucks.
I love the idea of the Alpha school. But do wonder how much of the success is self selective and probably good teachers. My kids use ixl and it is just a glorified quiz. Khan Academy is good but literally recordings of Khan lecturing.... not AI.
In cae you are wondering, Trump is the reason drones are falling on Abu Dabi.
Not surprised by the problem of the SF Bay Area data centers getting grid power. Place overall is notoriously left-leaning, overly concerned about the environment. Believe the small nuclear power plants will have a tremendous market. We don't have Waymos here, doubt I would ever use one unless I was no longer able to drive. Far as AI goes, not only do I avoid AI but don't use any 'smart' appliances unless they need not be hooked up to the Internet. Simply do not trust AI and 'smart' things in general. Have the background, was trying out VR glasses nearly forty years ago, was acquisition manager for two, later four large-scale computer-based training systems. AI is just one more tool, will never be a panacea, again, saw and heard all this stuff nearly forty years ago. Don't forget we need to put everything in perspective. Putting data centers in space is nice, but encryption of data streams going to and from the data centers will be paramount, and any encryption can be broken, just a matter of resources.
You recommend playing around with AI learning tools, do you have some recommendation with links that we can use to get started? I really enjoy your newsletter, I learn something new from every issue. Keep them coming.