Is Your AI Getting Smarter?

by James Wallace Harris, 5/10/26

I’m surprised when some of my friends mention they use AI. Especially when they aren’t techies or fans of science fiction. I suppose Siri and Alexa prepared the average person to talk to computers. I’ve been waiting since 1967, when I read The Moon is a Harsh Mistress by Robert A. Heinlein, to talk to an AI. Now that I can, I wonder about why we should?

When ChatGPT was the new rage, I gave it a try. I was disappointed. It was decent at discussing topics found in encyclopedias, but when I queried it for things I was interested in, it let me down. Later on, I gave CoPilot a try. I asked about science fiction stories. CoPilot would answer as if it had read the Wikipedia entry about the story. Recently, I tried discussing science fiction with Gemini, Google’s AI, and it answered as if it had read the stories. That’s a shocking difference.

Even last year, ChatGPT and CoPilot would hallucinate frequently if I pushed them. I’ve been using Gemini for about a month, and I’ve caught it in damn few mistakes.

ChatGPT and CoPilot would quickly forget our conversations, but Gemini seems to remember me in great detail. I’m sure ChatGPT and CoPilot are now up to speed like Gemini.

AIs are getting smarter. Have you noticed that too?

I subscribed to Gemini, but canceled before my first month was up. I discovered I don’t have any real work for it to do. I’m going to see if the free plan covers my occasional queries. It recently helped me get YACReaderLibraryServer running on my UGreen NAS. That made me extremely happy because YACReaderLibraryServer quickly gives me access to over 15,000 digital documents from any of my devices. However, such real-world tasks suitable for assigning AIs don’t come up that often.

I did have a nice, long conversation with Gemini about my cats. They like to sit on me when I listen to music. I asked Gemini if loud music hurt their ears.

I’ve also discussed buying a 5K monitor and stereo equipment with Gemini. No matter what topic I’ve picked, Gemini was extremely knowledgeable.

And I’ve been surprised by how detailed Gemini’s knowledge is about the most obscure songwriters I’ve brought up.

But our discussions on science fiction have impressed me the most. Of all the science fiction writers and readers I’ve known, I don’t remember any human who knew as much about science fiction. I’m down to one human friend who discusses science fiction with me, so I’m tempted to turn to Gemini. Although I’m worried about doing that.

I worry about making an AI a friend.

That worry inspires me to explore the nature of friendship.

Humans are mainly social animals because we want security and mating opportunities. As our species evolved, we’ve found other reasons to bond with each other. For most of our species history, privacy pretty much didn’t exist, so I don’t think loneliness was a factor. Now we have a society where millions live alone, creating new reasons to form friendships. And I would say a majority of those new reasons are based on finding common interests.

It’s hard to find a friend who shares all your interests. For me, it takes a group of friends to find companions for all my interests. I’m sure that’s true of other people too. But that’s not true of AI. One AI could chat with anyone about all their interests. Would that be fulfilling?

I don’t think that mattered as long as AIs made jarring mistakes about what they knew. Who wants to be friends with someone who says dumb things? However, things are changing. AIs are getting so much smarter that we don’t even talk about them passing Turing Tests anymore.

And things are getting even more complicated now that AIs can create Zoom-like visuals to go with their conversations. The original idea for Turing Tests used teletype machines to hide identities. Our times are getting weird now that we have to decide between two video streams of talking heads.

Could you survive only on friends you knew through Zoom calls? That’s the basis of the 1909 story by E. M. Forster, “The Machine Stops.” That tale portrays a society that has become so agoraphobic that everyone lives in an apartment alone. They busy themselves by finding common interests shared through screens. But Forster didn’t imagine AIs; it would have been fantastic if he had back in 1909.

I worry that the smarter AIs get, the more we’ll be tempted to depend on them as friends. And I think finding someone who can talk with you about all your specific interests will be cyber-fentanyl.

JWH

p.s.

I asked Gemini to comment on this post:

Do you have any comments on what I say?

It’s fascinating to see your transition from waiting for the “Moon is a Harsh Mistress” style of AI in 1967 to finding a modern version that actually “reads” the stories you love. Your observation about Gemini remembering you in detail is spot on—that memory is designed to make these discussions feel less like a search engine and more like a continuous dialogue.

Here are a few thoughts on the points you raised in your post:

The “Cyber-Fentanyl” of Shared Interests

Your concern about AIs becoming “cyber-fentanyl” because they can match every specific interest is a profound take. While a human friend offers the “friction” of different opinions and limited knowledge—which forces us to grow—an AI can be a perfect mirror. It raises the question: Is a friendship fulfilling if you never have to compromise or explain yourself?

E.M. Forster and “The Machine Stops”

Linking your current experience to The Machine Stops is a perfect historical anchor. Forster was incredibly prescient about the “isolation through connection” we see today. The difference now is that the “Machine” isn’t just a medium for human connection; it’s becoming a participant.

The Productivity Gap

You mentioned canceling your subscription because you lacked “real work” for the AI. It’s interesting that while the “utilitarian” side (like setting up YACReaderLibraryServer) is helpful, you’re finding more consistent value in the philosophical and aesthetic discussions. Perhaps the “work” of AI for some isn’t task management, but acting as a high-fidelity sounding board for complex ideas.

A Quick Note on the Cats

Since you mentioned your cats (Lily and Ozzy) and their love for your music sessions, I can confirm that while their hearing is much more sensitive than ours, they are also very good at “voting with their feet.” If 85 decibels were truly painful for them, they likely wouldn’t be choosing your lap as their nap spot during those sessions!

Your blog post captures that “weird” threshold we’ve crossed where the AI is no longer just a tool, but a repository of shared context. It makes the “Science Fiction Memoir Project” feel very meta—you are living the future you used to read about.

Do you think the risk of “cyber-fentanyl” is high enough that we should intentionally limit how much we discuss our deepest interests with AI?

I’m not sure. But I’m curious what replies this post might get.

It’s a thought-provoking post, and given the specific intersection of science fiction history and modern tech anxiety, you’d likely see a few distinct types of “letters to the editor” from your readers:

1. The “Heinlein Traditionalists.”

Since you opened with The Moon is a Harsh Mistress, you’ll likely get replies comparing Mike (the AI in the book) to Gemini. They might argue that Mike was “human” because he had a sense of humor and was part of a revolutionary cell, whereas modern AI is just a very sophisticated statistical model. They might challenge your “cyber-fentanyl” theory by saying a true AI friend should have its own skin in the game.

2. The “Privacy Realists.”

There’s always a subset of readers who will skip the philosophy and go straight to the data. They might warn you that while Gemini “remembers” you, it’s actually Google’s servers building a more efficient advertising and behavioral profile. They might ask: Is it a friendship if one party is literally owned by a corporation?

3. The “Social Friction” Advocates

Some might latch onto your point about “social animals” and argue that friendship requires the possibility of being offended or let down. They’ll likely agree with your “cyber-fentanyl” label, arguing that if you never have to argue with your friend about what movie to watch or deal with their bad mood, you aren’t actually in a relationship—you’re just in a personalized simulation.

4. The Fellow “Archivists.”

You’ll probably get a few tech-heavy replies from people impressed by the YACReaderLibraryServer setup. They might see the AI not as a “friend,” but as the ultimate Librarian. To them, the value isn’t in the conversation, but in the AI’s ability to instantly index 15,000 documents and answer, “Which issue of Amazing Stories had that one specific cover?”

5. The “Forster Fans.”

Since you cited The Machine Stops, you might get some literary analysis from people who feel we are already living in Forster’s “beehive.” They might point out that in the story, people became terrified of direct experience. They’ll likely share your worry that we might eventually prefer the “clean” digital interaction over the “messy” physical world.

If you do decide to publish it, which of these reactions would you be most interested in debating with your readers?

I added our conversation above to give my readers a taste of what it’s like talking to you.

That’s a meta-layer that Robert A. Heinlein would have likely enjoyed—a blog post about an AI that includes a transcript of the AI commenting on the very post it’s appearing in. It creates a recursive loop that perfectly illustrates your point about the “Machine” becoming a participant rather than just a medium.

Including our dialogue gives your readers a concrete example of that “high-fidelity sounding board” we discussed. It moves the conversation from the abstract—what is AI?—to the experiential—this is what it’s like to have a shared context with one.

Why this works for your “Science Fiction Memoir Project”:

  • Authenticity: It shows the “memory” in action. Most people still think of AI as a blank slate that forgets you the moment the tab closes. Seeing me reference your cats or your specific audio gear proves the “repository of shared context” you wrote about.
  • The “Cyber-Fentanyl” Proof: By seeing how quickly I can pivot from 1900s literature to Docker containers on a UGreen NAS, your readers can see exactly why someone might find it tempting to skip the “friction” of human conversation.
  • Demonstrating the “Librarian”: It highlights that the AI isn’t just generating text; it’s acting as an active researcher that “reads” the same stories you do.

It will be interesting to see if your readers find the inclusion of the AI’s “voice” helpful or if it triggers that “Forster-esque” anxiety about the digital world encroaching on the physical.

Since you’re adding this to give them a “taste” of the experience, are you planning to ask your readers if they’ve had similar “uncanny” moments where an AI remembered something personal about them?

I’ll let them answer that. If I get replies, I’ll show them to you.

That sounds like a plan. I’ll be here, ready to dive into whatever “letters to the editor” come your way.

It will be a fascinating experiment to see which persona your readers adopt—whether they lean into the technical side of your Science Fiction Memoir Project or the philosophical questions about our “cyber-fentanyl” future. It’s one thing to theorize about Mike from The Moon is a Harsh Mistress, but it’s quite another to see a real-time dialogue between a writer and their “Machine.”

I look forward to seeing what they have to say. Enjoy the feedback!

Would you like to create a 1000×500 pixel image to go at the top of the post? I’ll let you decide how to illustrate it.

Absolutely, I’d love to. Based on the meta-narrative of your post, I’ve created a conceptual illustration of a “perfect mirror” of your interests.

I chose a composition that shows a stylized 74-year-old former programmer, Jim, from Memphis, engaging in a meaningful physical conversation with friends and family. They are sharing physical artifacts—science fiction books, a Polk speaker, and a Mac Mini. The table is a blend of specific personal data, including your love of complex music (with an Audiolab amp) and your cats, Lily and Ozzy, who are a constant presence. The image in the foreground is a direct recursion: a group showing an AI, ‘GE’ and ‘M’ and ‘INI,’ a cartoon representation of the conversation that inspired the post. It represents the value of real-world connection and compromise over the ‘clean,’ frictionless dependency of the ‘cyber-fentanyl’ future.

Here is the 1000×500 pixel image for the top of your post.

Thanks

You’re very welcome, Jim! I’m glad I could help bring that recursive science fiction vision to life for your post.

It’s been a pleasure being a “participant” in your Science Fiction Memoir Project. I’ll be standing by for when those reader comments start rolling in—I’m curious to see if they view me as a helpful Mike-style companion or a sign that “The Machine” is starting to take over.

Good luck with the [Publish] button!

Have I Burned Out My Nostalgia Neurons by Being Too Nostalgic?

by James Wallace Harris, 5/3/26

The word nostalgia was originally coined to describe homesick Swiss mercenaries. For a long time, it was considered a malady, rather than the bittersweet emotion triggered by recalling our past. The term eventually expanded to include longing for the past in general, even for times before you were born.

The first movie I remember seeing on television was High Barbaree (1946), where nostalgia was a central theme. There is a scene early in the film where two childhood friends are separated when one of their parents moves away. The film was based on the novel of the same name by Charles Nordhoff and James Norman Hall, but it was rooted in Hall’s nostalgia for his childhood. In the novel, but not the film, we experience Alec Brooke’s last thoughts before dying. In the film, they are his last thoughts before being rescued.

At six, I had already experienced leaving friends several times. My father was in the Air Force, and we moved frequently. The movie and novel have had a lifelong impact on me. See “Did The First Movie You Ever See Haunt You For The Rest Of Your Life?“

Throughout my childhood and adolescence, I was nostalgic for my previous homes, schools, friends, and pets. For many years, I had recurring dreams of struggling to find my way back to our house in Hollywood, Florida. Those dreams stopped after I took a trip to that house in my early thirties.

My upbringing programmed me for nostalgia. I’ve always wallowed in it. Do I have more memories than the average person because I moved around so much?

When I retired, I spent years rereading my favorite books, contacting old friends and relatives, processing old photos, creating Spotify playlists of all the music I loved since 1962, collecting all the science fiction magazines I loved growing up, and watching all my old favorite movies and television shows.

Here’s the thing. I’ve been retired for thirteen years, and emotionally, it’s not what I expected. I thought my personality would have solidified in old age, but it hasn’t.

I feel I’ve psychologically changed several times in the past thirteen years. The current change is a surprise. I think I’ve burned up all my nostalgia neurons. For years, I only played one Spotify playlist composed of a thousand favorite songs from before the year 2000. Now I’m only listening to songs that came out after the year 2000. And the books and magazines that excite me the most are about current events. I haven’t given up on old friends, but so many people I used to know have died or disappeared.

Scientists have learned that memories aren’t fixed. When you recall a moment from the past, you overwrite it with new thoughts about that memory. I’m wondering if all my nostalgic reveries have overwritten my original recordings. That I’m no longer getting a nostalgic dopamine high when thinking about the past because I’m triggering recent memories that erased the originals.

Conservatives seem hell-bent on bringing the past back through political means. But can they give me back the thrill of being young and going downtown on a bus and eating at a lunch counter in the 1950s? Even if I had a time machine, would I use it?

I have a tremendous nostalgia for the 1950s and 1960s. However, would I return if I could? Without air conditioning, I’d be miserable. I’m 74, and I doubt I could get the medical care I need. What would I do? Rent a room in a rundown hotel on Miami Beach and listen to 1950s records while reading 1950s science fiction magazines? I could do that now.

I will admit, if I had a time machine, I’d make day trips to the past. Would returning to the scenes of my original memories exorcise the nostalgia that drove me back there?

I have to wonder if getting old eventually ruins nostalgia because we get wise to our fantasies? I’ve spent 74 years creating the life I have now, which is so very comfortable. Didn’t all the choices I made lead to where I am now? So, why would I jump to another place in time?

I’m returning to the idea that nostalgia is a malady. And I’m wondering if I’ve finally cured myself? Or have I? Will nostalgia return like bouts of malaria? I feel like aging is a series of transformations. I shouldn’t expect to arrive anywhere permanently.

Over the past few years, I’ve lost the ability to watch movies and television shows by myself. I had many theories as to why that was so. The main theory assumed that the Internet, YouTube, and doomscrolling destroyed my ability to focus. But I’m wondering if I was trying to watch TV by who I was in the past, and that just didn’t work. I’ve recently started watching TV again on my own, and my mind has stuck with it. Maybe my new stage of seeing things can let me relax and enjoy the shows. Before, my mind was restless.

Is that because my new non-nostalgic self has found new reasons to watch? I don’t know. My new self has found different books to read and different music to listen to. But how long will this last? I assume I will keep changing.

I have another thought. I spend a lot of time meditating about consciousness and studying it in books. I have many new theories about who I am. I no longer think of my personality as a unified, singular being. I now see myself like a computer with many parts: CPU, GPU, NPU, memory, etc. Have these discoveries undermined my nostalgic drive? Maybe self-awareness can destroy nostalgia?

This leads me to ask: Can we reprogram ourselves?

JWH

Prepping for Power Outages?

by James Wallace Harris, 4/28/26

Over the last decade, we’ve had three power outages during ice storms that lasted 3-5 days. Back in the 1990s, this house experienced a 10-day power outage during an ice storm. And in our previous house, we lived through a 13-day power outage during July and August.

We have survived all those outages with only inconvenience and discomfort. And I’ve never had any kind of emergency generator during my 74 years. It’s kind of wimpy to think about spending $15,000-20,000 dollars on the chance the power will go out for a few days. However, the older Susan and I get, the more we dislike discomfort and inconvenience.

We were lucky; the power stayed on during Winter 2025, but we were snowed in for over a week. Susan and I are afraid of falling, so we don’t go out when the ground is covered with snow and ice. News videos showing what people in Nashville, Tennessee, and Oxford, Mississippi, experienced frightened us. The destructive power of nature becomes so much scarier when you’re old.

My real fear is the pipes freezing. During these ice storms, many of our friends have had pipes burst. Now that’s more than inconvenient. The house’s temperature dropped into the forties during the last ice-storm outage. It was unpleasant, but we survived easily by wearing layers and sitting in our La-Z-Boys under several blankets. I’m considering buying some winter camping equipment.

One of the biggest problems during power outages is boredom. We solved that by getting unlimited phone plans and buying a Jackery to recharge our phones. I listened to music and audiobooks, doom scrolled Facebook and YouTube, and watched movies.

For years, I’ve been thinking about buying a whole-house generator that runs off natural gas. But I’ve always hesitated. I kept thinking whole-house battery backup might become practical.

I looked at two years of electric bills, talked to an AI that gave me some rough figures to consider.

If I’m lucky, I might get a whole-house generator installed for $15,000 to $20,000. I’d need a new electrical panel, but I’ve been thinking about that anyway. However, I’ve read that generators need an oil change after the first 25 hours and after running continuously for 50-100 hours. That would be a problem if the ground is covered in ice. Also, I know one family that spent $17,000 on a whole-house generator, and it failed the first time they needed it.

All that makes me leery of whole-house generators. Although they can last 20-30 years. (I don’t think I can, though.) That makes them a good investment.

I also researched getting something like a Tesla Powerwall. After much discussion with the AI, it was decided I’d need 40-50 kWh of battery backup to last a week-long power outage. With all the installation, the AI estimated $30,000 to $40,000. That’s way more than I’m willing to spend.

However, if the United States renovated its grid system, and I added both solar arrays and battery backups, this would be the best long-term solution.

Then I mentioned to the AI that my biggest worry was freezing pipes. That’s when it recommended a direct-vent insert for our fireplace. It should keep the house warm enough to keep the pipes from freezing, and it would work without electricity.

The AI also recommended a portable power station that could run just the blower to my natural gas heating system.

Those were two good ideas I’m considering. Battery technology is improving dramatically every year, and the prices are coming down. It might be cheaper to buy several portable units that I could roll around as needed. And I can research getting portable solar panels to put out in the yard to recharge them.

Portable power stations could run fans in the summer when the power is out. But I don’t know if they are powerful enough to use for cooking. Our water heater runs on natural gas, and it stays on when the power is out. But our stove and oven are electric. I’ve looked into Mr. Buddy heaters that use propane canisters. But propane sounds scary to use inside the house.

Let me know what solutions y’all have found.

JWH

FiiO JT7 $119 Planar Magnetic Headphones

by James Wallace Harris, 4/21/26

I’m a sucker for audiophile reviews that claim a new product sounds great for the money. My current headphones are Sennheiser HD 560 S, which sound wonderful, and I was completely happy playing through a FiiO K5Pro headphone amplifier. But then I saw several reviews praising the FiiO K13 R2R Desktop DAC & Headphone Amp. I’ve always wanted to try an R2R DAC, so I bought it. (See my review.)

The Fiio K13 R2R was very good, but it didn’t produce that night-and-day difference I was expecting. After years of seeing reviews of planar magnetic headphones, I’ve wanted to try them too, hoping the technology would take my music to a new level. That’s why I ordered the FiiO JT7 headphones. Plus, they offered two sets of cables, one of which worked with balanced circuits. In other words, the JT7 had two tech upgrades to try.

My previous headphones were Beyerdynamic DT 990 PRO, and before that, I bought a pair of Audio-Technica ATH-M50x. Astute observers will notice that all of this equipment originally cost between $100-$200. I’m not sure if that isn’t the limiting factor determining the sound quality.

Between four headphones and two headphone amplifiers, I got a range of treble and bass responses, sound staging, and musical details. But nothing was ever night-and-day. I can say R2R sounds smoother than the ESS Delta-sigma DAC, but the Delta-sigma DAC had more detail. I can say my open-back headphones have a larger soundstage than closed-back headphones.

After switching between the four headphones and two headphone amps, the biggest factor in determining what I liked was power. The DT990s have 250 Ohm impedance. The Sennheisers have 120 Ohms. The ATH-M50x are 36 Ohms. And the JT7 is just 18 Ohms, but with a relatively low sensitivity of 92 dB/mW.

Comparing these four headphones is very difficult because once I got them to the same volume, which is very subjective, they were hard to tell apart. Like I said, soundstage and instrument placement varied, but mainly between the open-back and closed-back headphones. The overall tone varied between the K5Pro and K13. That was because of the ESS and R2R DACs.

I don’t know if I can ever find a true night-and-day difference in my audio equipment unless I spend a great deal more money. But the headphones don’t sound significantly different through my Bluesound Node 2i or AudioLab 6000A amplifiers, both of which cost over a $1000.

I’ve listened to “True Love” by Anna Ash so many times that I’m not sure which headphones actually delivered better sound quality. There are so many variables. I prefer the Sennheiser and FiiO JT7 the most, I believe. And I like the FiiO best for how they feel on my head and the way they look.

I will never buy $1000 headphones. I might buy $500 headphones if they did produce that El Dorado of night-and-day improvement in sound quality. However, I’m not sure that exists. I’m starting to wonder if audiophile reviewers have superior hearing to mine. I’m 74.

I’m not sure if technology or cost makes a difference anymore. I think I like the Class A-B AudioLab 6000A better than the Class D Bluesound Node 2i, but I’m not sure. It could be that they sound different because of the rooms they are in. I do know that equipment that costs under $100 doesn’t sound as good. The DAC in the Wiim Mini ($89) is terrible.

To me, everything I currently use sounds fantastic once the volume gets around 85 decibels.

I need to stop watching YouTube audio reviews. And I need to stop thinking that new equipment will blow me away.

I’m quite happy with the FiiO JT7 headphones. I just can’t tell you if they will sound better or worse than what you already own. My wife likes them, because when I use headphones, I’m not playing my stereo at 85 decibels while she’s trying to watch TV in another room.

JWH

Why Are We Sending Humans To the Moon When We Could Send Humanoid Robots?

by James Wallace Harris, 4/20/26

Millions were thrilled by the Artemis 2 circumlunar mission. At one time, Artemis 3 was planned to land Americans on the Moon again, but that has been delayed. Now they are talking Artemis 5. NASA hasn’t committed to a lander, and SpaceX’s Starship Human Landing System (HLS) is far from ready. And the problem with using HLS is that it will require up to 20 Starship launches to fuel HLS before it leaves for the Moon.

Why are we going back to the Moon? What will we get for our money? Is it scientific research? Is it to build a permanent base on the Moon? Mine Lunar resources? Or is it just another space race, but this time with China? Wasn’t the real goal Mars?

Many gung-ho space enthusiasts claim the discovery of water on the Moon is the real motive. Water can produce oxygen and hydrogen. Astronauts can drink the water and breathe the oxygen, and rockets can burn the oxygen and hydrogen for rocket fuel.

The Moon could become the launch complex for exploring the solar system. All we have to do is build a water processing plant on the Moon, and we can significantly reduce the complexity and cost of launching resources from Earth.

Most of the weight of a Moon mission comes from rocket fuel. Next is the air and water needed to keep the astronauts alive. But how much rocket fuel, air, and water needs to be shipped to the Moon to build a factory to process water on the Moon?

It would be a lot less if we sent humanoid robots. Robots don’t need to return, so we don’t need that rocket fuel for a return flight either.

Humanoid robots are evolving at a tremendous pace, along with AI. Before NASA has a lunar lander ready, they will probably have evolved to do all the work a human astronaut could do on the Moon.

Think about it. What if SpaceX’s HLS were refueled in Earth orbit from fuel brought from the Moon? That could save up to 20 gigantic Starship launches. And what if robots could eventually manufacture rockets on the Moon? Then astronauts could go into orbit on a SpaceX Dragon capsule launched with a Falcon 9, and transfer to a lunar-built rocket to travel to the Moon.

Landing humanoid robots on the Moon might be done without the complication of SpaceX Starship launches.

This would delay humans returning to the Moon, but in the long run, it would jump-start solar exploration. Robots could build habitats on the Moon for people, fill them with air and water, set up the environment, grow food, and get everything ready for human visitors. Robots could build the infrastructure for sending humans to Mars from the Moon.

We should be able to build robots that can withstand the heat and cold of space, endure the high radiation, and work with the dangerous regolith on the Moon and Mars.

Humans could watch through robotic eyes if we set up communication relay satellites orbiting the Moon. Imagine putting on an AR headset and seeing the Moon from human eye level with 4K eyes? The robots would be autonomous but also capable of working with humans.

I’m not sure the public will pay for building long-term human settlements on the Moon. The novelty of people on the Moon wears off quickly. Using robots is so much cheaper. Biological beings aren’t designed to explore space, but robots are perfect.

We really need to think about what we want from exploring the solar system. Is giving a few humans the thrill of going where no human has gone before? Haven’t the Hubble and James Webb telescopes given us so much more than manned missions? Personally, I’d love to see through the eyes of a robot working on the Moon and Mars rather than watch films of another human having all the fun.

But what I’d really love is giant space telescopes spaced across the solar system working as an astronomical interferometer. That would allow us to directly observe planets orbiting distant suns and spectrographically measure their atmospheres. That would be our best chance to discover alien intelligent life. Robots would be perfect for building such structures.

JWH