214: Analog Computers – Old is New Again

On today’s episode of Still to be Determined, we’re going to talk about how Sean has a cold. No, no, we’re not. We’re going to talk about analog computers and how they’re making a comeback. In fact, we might even talk about, did they ever leave? That’s right. We’re going to talk about Matt’s most recent episode, why the future of AI and computers will be analog, which dropped on April 9th, 2024.

As usual, I’m Sean Ferrell. I’m a writer. I write some sci fi, write some stuff for kids, and I’m just generally curious about technology. Luckily for me, my brother is that Matt of Undecided with Matt Ferrell, which takes a look at emerging tech and its impact on our lives. And with me as always is the aforementioned Matt.

Before I mention Matt, how are you?

I’m good. How are you doing other than having this cold that’s like, um, kind of rearing its ugly head?

This cold is, uh, really kind of a souvenir from France. I recently took a trip with my partner. We took a Vacation, the first in, I don’t even know how many years, six years, seven years predating, well predating the pandemic.

And so we went to France, we had a lovely time and on the plane ride home, just as they were lowering the wheels on the plane to say, and now we’re back in New York. I was like, is that a tickle in my throat? Am I, am I overly aware of swallowing right now? And I told Matt before we started recording, I am now at that stage where I probably sound worse than I feel because I feel pretty good.

Uh, and luckily for me, and more importantly for everybody on YouTube who’s watching us right now, We didn’t record this on Wednesday when I looked like Jabba the Hut if he had been turned inside out. It was nightmarish.

That’s a

beautiful

picture you

just painted,

Sean.

Oh, yes. I am a wordsmith. I’m a writer.

I’m a writer. Uh, I may have mentioned that before. Um, yeah, so it was pretty terrible. So I may sound a little froggy right now. But I do feel okay. And I’m looking forward to this conversation because this is the kind of technology that 12 year old Sean was just like, how do we build it? How do we use it?

What can we do with it? I was, we, I don’t know if you remember this Matt, we had a small electronics game that was basically an engineering analog computer style thing. with wiring and circuitry. And you were supposed to be able to put it together and do things like create a microphone with a speaker or rewire it so that it would become a waveform simulator and all those sorts of things.

And all we could ever really do that I remember is make it make horrible hissing sounds that would cause our father to come running out of his den and say, you got to knock that off, go outside.

But that was effectively, that was effectively a analog Computer. That is what we were playing with. And so this kind of tech, I won’t say returning, just sort of re emerging. I believe that this kind of thing has been lurking in the background for a long time. And I found some comments on your video that do reflect that.

Um, so I’m looking forward to talking about it, but before we get into the conversation, what was it about this that kind of pinged your radar and made you think, Oh, this is something that should be talking about on the channel.

Honestly, it was one of those. I knew this was always bubbling around the background.

But there was a video from, I mentioned this in the video, uh, from Veritasium from almost two years ago, I think, where he put a thing about analog computers and talked about a company called Mythic. And for me, it was kind of like a, Oh my God, I didn’t know that was a thing. I didn’t think there was like this reemergence of how can we reapply analog computing to today.

And his video kind of talked about basically like how. And all computers kind of work. And I thought there’s a different way we could kind of look at this. And so I started digging into it and got more and more interested, kept reading more, brought it up as an idea to my team. And we started like really going deep into it.

And my lead researcher kind of went down a rabbit hole. It was, it was like an epic, an epic journey that Vincent went on. That was fantastic. Uh, but yeah, that, that was kind of the, the, the trigger point was seeing that other video from another YouTuber. talking about this interesting company and analog computing that made me go, Whoa, where can we use this?

How can this like evolve? Why did it go away? Why is it coming back? All those kinds of questions that were bubbling around my head are why we did this video.

Very cool. I’m looking forward to our chat about that. Before we talk about Matt’s most recent, I wanted to revisit some comments from our earlier episode.

This is from episode 212 in which Matt and I discussed sand batteries. It’s getting hot. And there were comments like this from Pixel Pusher, who wrote in to say regarding solar energy payback methods, it seems like the industry missed one great marketing concept. Use 100 percent solar energy for the production of future panels.

That way there’s literally zero time required to offset the energy used to create a panel. I wanted to Throw that your way because my understanding is it was never quite that easy and might reach that point at some point in the future, conceivably now, but when this industry all started, there wasn’t enough solar energy being produced to actually do this.

Correct?

Correct. But it’s also a really thorny kind of like the way you have to think about it. There’s the mining, you have to get materials, you have to get aluminum, you have to get all that kind of stuff that is putting out emissions and takes energy, which is mainly from fossil fuels, from the machinery and all that kind of stuff.

Then there’s the actual factory that’s making the stuff, which is maybe not coming from clean energy, which is the point that they’re raising, which is, yeah, if you power that factory by solar energy. It’s helping. I think a good analogy would be like electric vehicles. Like one of the arguments against EVs is, Oh, great.

You have this car that’s not putting out emissions, but to charge it, you’re still charging it off a coal power plant. Like, and that may be true in certain regions of the world and the United States, but guess what? The grid is getting cleaner and cleaner and cleaner every year. So over time, EVs are going to get cleaner and cleaner and cleaner to charge, which you can’t say for fossil fuels.

It’s the same thing for this. So solar panels are, recycling is starting to kind of take hold, uh, heavy industry is moving towards cleaner technologies like hydrogen powered. mining machinery, even EV, like a battery powered machinery. So all of these things that equate to why it’s not clean to make a solar panel and what, how much energy it takes is getting better over time.

But I have no idea how long it’s going to take to get to that hundred percent cleanliness. Right. It’s probably by like 2050. You know what I mean? That’s probably like a ways off before that ever happens.

Right. This is all to say the pixel pushers, uh, goals are a good one. Yes. Like that as a goal, 100%, 100 percent purity or as close as possible is always the goal.

Um, but that doesn’t mean that right now there are very good arguments for leaning more heavily in that direction, even though we’re not there yet. There was also this comment from Sarah Loy who was taking the argument about as simple or as strange as it sounds to say, heat up the sand, store the sand, reclaim the heat later from the sand.

She jumps into the comments to point out that this is actually something that’s been going on for a long time in farming. Greenhouses, she says, are being built now with heat storage in the soil underneath. All summer the circulation brings heat into the soil. and cooler air into the greenhouse to prevent overheating.

All winter it does the opposite. There are some large examples in Canada. I thought that that was a really fascinating use case, and I wondered if that was something that had come across your radar in your research.

Not for this specific video, but that comment does make me realize Also in the video, we probably should have brought stuff like that up as well as passive homes because passive homes operate the same way.

It’s like you orient the house and you have windows and then you have like, uh, large amounts of mass, like cement floors that would absorb the solar radiation during the day. And then it’s slowly radiating that heat off overnight. And then in the winter, in the summertime, when the sun’s higher in the sky, less of that solar energy is reaching inside the house to heat the cement.

So it’s not going to have that same effect. You’re taking all this stuff into account to try to store that heat energy in a way in a passive process, which is the same example of the greenhouse, which is really cool. We should have brought that up. A hundred percent.

Finally, I thought this comment was a great summary of our conversation, which we’ve had before.

I’m sure we’ll have again about critics of sustainable tech and the arguments against the negative effects of it. Ryu Kagedesu says the argument over the negative effects of sustainable tech reminds me of the old story about a group cutting a path through the jungle. Men cutting the brush, men keeping the tools sharp, men cooking to keep everyone fed.

One guy climbs a tree and shouts down, Hey, wrong way. Someone yells back, Who cares? We’re making progress. I thought that that was a very amusing twist on like when somebody is yelling, you’re going the wrong way. It doesn’t matter how quick and efficient your work is going. Maybe you got to stop, take a breath and say, all right, let’s reassess.

What are we doing? What are our goals and where are we headed? So thank you for that Ryu. I think that that was a really great little joke to reframe the conversation. On now to our conversation about Matt’s most recent, why the future of AI and computers will be analog. There were a number of people who showed up, apparently some of them out of retirement in your comments, just to say I used analog computers for the.

Large portion of my career, they’re like this from Bruce Fay, who said, I studied analog computers and computing in the 1970s as part of my electrical engineering education. At one time after that, I worked for a company in Ann Arbor, Michigan that made some of the most powerful analog computers in the world.

I was in marketing at that point. They were used among other things to model nuclear reactors and power plants. Incredibly powerful. I want to stop for a moment and just. Marvel at the fact that they were used to model nuclear reactors. This is not, oh, you’ve got this little device that allows you to pull a lever and it spits out a little punch card at the other end that maybe helps you speed up the process.

Accounting issues. These are very powerful computers. They are tasked maybe to a more narrow scope of work, if my understanding is correct. So you have this device, which as opposed to a Windows machine or a Mac, where I’m able to sit down and do whatever I want with word processing, web design, gaming. I can, you know, mix and match my day to do anything that I care to.

On the, on a PC, but just because something is maybe gears and circuits that are lined up with one task doesn’t mean that that one task has to be a simple one. Modeling nuclear reactors and power plants. Yeah. The complexity there and the sophistication that’s needed for that. Do you want to talk a little bit about, he mentions The research he was doing and the work he was doing in the 1970s is part of his degree.

Do you want to talk about some of the work that was done by these computers during those decades prior to the digital era jumping forward? You did talk about the space age. We got to the moon on less computing power than we currently carry in our pockets. But were there other places that your research pointed to, to powerful analog computing that might stun a viewer or listener today?

I don’t know if it would, I don’t think it’s going to stun, but it shows you how far back this goes. I mean, I mean, analog computers go back decades. Centuries and centuries and thousands of years, abacuses or analog computers, but it’s, uh, somebody on X actually reached out to me with this. And we did kind of come across this in not this specific example, but examples that went back into the, like the forties and the fifties.

You can think about World War II. We were using analog computers all over the place in World War II. And one of them was on our guns, on our planes. Uh, there were analog computers that helped the targeting system. As you were aiming your gun for how fast the plane was going, how fast the bullet goes, the arc and the trajectory of the, of the bullet to make the, uh, sight line up in a way that the person aiming the gun would actually be able to hit their target.

Because there’s so many variables happening at once, it’s really hard to do that. And they had analog computers on these guns and the targeting systems to help. Make those hits. Same thing with the bombs that we were dropping. The, the, the, the bomb mechanism that the bombers looked through was basically an hill log computer and it was so proprietary and so like, we cannot let the Germans get ahold of this if they, on our planes.

It was primary of importance that if you were getting shot down, like if the bomber was going down, you had to take that thing with you or destroy it before you jumped off the plane. So it was. These analog computers were in war in the 40s. They’ve been part of massive things like space flight and figuring out trajectories and orbits and things like that.

They’ve been used for countless number of examples of things because they’re perfect for, as we brought up in the video, differential equations. Yeah, that’s so much of stuff in engineering and physics and all those kinds of things that we’re doing that that makes perfect sense, which is why analog computers, as you pointed out, are very, very

tailor made. If you have a specific thing you’re trying to figure out, and it’s a repeatable thing that you need to keep calculating, figure out. Analog computer. There you go. You can just custom make that analog computer to do that specific task and calculation, and then it speeds up that process again and again and again.

And Like I said, there’s countless examples over history, there’s too many to go through. But for me, I thought the one from that person that commented to me on X was a good one of the World War II example.

Yeah, it reminds me of, I mean, another example from World War II would be the Enigma machine. And the work by Turing that was the Turing who is viewed at now as, you know, the great grandfather of modern computing and his computing was all analog.

It was the break, the code breaking, the code making were done by machines that would have been analog computers. I’m also reminded, I can’t remember what they call it, but there is a device that was discovered in the Mediterranean. It is a ancient, uh, They don’t know whose it was, but it was an ancient device, which they don’t even know what the purpose of it was.

But it is a bunch of gears locked in together with each other that has to be some sort of analog computer that was possibly used for navigation. And you mentioned, um, you know, old centuries, old examples like the Abacus, old seafaring devices used for measuring angles to stars and being able to chart out your path would have been a form of analog computer.

So all of that, it does really force us to take a step back and reassess what we, what we mean by computer. And I think that our modern era, I am guilty of this. I know. I think if a computer is a thing I plug into the wall and it’s got a keyboard, it’s got a screen. That’s not the word compute literally refers to what you mentioned, differential equation calculations, number crunching.

There was also this comment in Matt’s video from Leg who wrote in to say, I built and used analog computers, although I didn’t call them that, to control the temperatures in a multi layered cryostat for my PhD work in the mid seventies. I did the number crunching on a digital computer that filled the basement of the math department building using punch card input.

Twenty odd years later, I found myself working with an engine management computer for a helicopter that was pure analog. When I approached the system engineer at a large well known aerospace company who had the design control authority for the system to ask some fundamental questions about it, he didn’t have a clue.

He was purely digital. I’m retired now, but if I drag my knowledge out of the closet with my flare jeans and tie dyed t shirts, maybe I’ll come back into fashion. Like, I don’t know that you ever went out of fashion. I think it’s one of those things where sometimes the things in our closet that we think are so out of style.

Are really just waiting for the right context and I think there is a certain amount of expertise in retirees that I wonder if the industry at large is aware of the repository of knowledge that might be available if they just opened up the call to say, Hey, people who studied engineering in the sixties and seventies, we’ve got some questions because the new applications might benefit from that wisdom.

What do you think about that?

Oh yeah. A hundred percent. It’s there’s, I do agree with the comment of like, what’s old, you know, it’s a cycle. What’s old becomes new again. What’s old becomes new again. And just, we keep repeating this over and over again. Uh, I mean, it kind of ties back to some of the other comments that you’ve already read through where it’s like, this stuff has been around forever, almost has never kind of gone out of style.

It’s always just been bubbling around in the background and we’re just not aware of it. It’s just, there are people that are using it, and then the rest of us are walking around thinking, oh, everything’s digital, but in reality, this stuff’s always been there, always has been, always will be. It’s just getting applied in ways now that’s becoming a little more, um, I don’t know.

front of, uh, front of mind becoming more a little more of in focus to the rest of us. Like that’s part of the reason why this appealed to me as a topic, which was, I didn’t understand that this was all bubbling around in the background for so long as it was. So there’s also a

question here about, uh, hybridization from Bill Miller.

The biggest drawback he says of analog circuits is that they are quite specific to a problem as we’ve mentioned in our conversation so far. So making something that will work for most or for generic problems is difficult. Where a digital computer is trivially easy in comparison. But when you need something specific, the analog computer can be significantly faster and more energy efficient.

I look forward to the hybrid components that will be coming out in the future. About the hybridization, Do you want to talk a little bit about that? What kinds of hybridization did you see in your research? And what do you think might be some places where you wouldn’t be surprised to see an emergence of hybridization between digital and analog?

This is not a sexy answer, but it’s a, this is why I think like from a consumer, it’s, it’s think about like internet of things. Think about, um, our phones and our smart speakers where you call out to some kind of digital assistant. They always have some kind of wake word where you say, Hey, blah, blah, blah.

And it responds to you. It’s doing that digitally right now. You could have that be done with analog, which would require far less power. And it would be just as accurate. It’d be accurate enough. It’d be great. So it’s like there’s ways that you could do this that would have those smart speakers, those little internet of, internet of things devices, smart watches.

You can create a hybrid approach where it’s still digital to do the bulk of the stuff it’s doing, but then there are these very targeted things that you have specific, specific little analog chip that’s made that’s in there. that handles wake words and things that are common questions and words that it would get that it needs to respond to in a way.

So it would make it much more energy efficient, make battery life longer, all those kind of things. Um, so that’s kind of, I think where we’re going to start seeing this stuff in our daily life. That seemed that that one seems pretty obvious to me, uh, where we probably see it. Uh, we looked into a company called Mythic.

It was the same one that was brought up by Veritasium in that video a couple years ago. Um, and what’s funny is after this video came out, the CEO of Mythic actually emailed me saying, I have been a subscriber to your channel for a while now and I was kind of surprised to see this video come out about his company.

Um, it’s, I, his, his company is doing exactly that. They have a chip where it can be kind of tailored in a specific way to handle certain kinds of things. And I could totally see, Computers or systems that have this mythic chip in combination with a standard CPU next to an NVIDIA GPU that’s doing all this crazy processing.

It’s not like this stuff is going to take out NVIDIA. It’s not like this stuff is going to make all of our digital stuff obsolete. It’s like, you’re going to see these systems just slowly get integrated. And I wouldn’t be surprised if at some point we see some Apple presentation where they always have the, you know, and this is the new M1 chip that’s on here.

And we created this new device that has the V2 chip that’s custom made for this. I would not be surprised if at some point, some of those systems that they’re talking about are analog because they’re going to handle specific things that can be tailor made like this.

I want to end the conversation. with what felt to me like a interesting synergy between two commenters, one being highbrow, one being lowbrow.

Let’s start with the highbrow. Sean, Sean H. Corey jumped in to say circa 1960 quote, it would appear that we have reached the limits of what is possible to achieve with computer technology. Although one should be careful with such statements as they tend to sound pretty silly in five years. This is a quote from John von Neumann.

John von Neumann was a Hungarian and American mathematician, physicist, computer scientist, engineer, and polymath. He worked on the Manhattan Project and he is, in his Wikipedia article, referred to as having, at the time of his work, receiving the widest coverage of any mathematician at his time. Balancing that out is this comment from Mike Shanahan who jumped in to say, my analog computer has 10 fingers and 10 toes.

Thank you, Mike.

Slow clap, Mike.

So there’s the yin and the yang of our conversation about analog computers. Listeners, viewers, what did you think? Let us know in the comments. Don’t forget, you can jump in there, leave a comment. That does help support the channel. You can also subscribe, leave a review, and share us with your friends.

All of those are great ways to support the channel and your conversations in the comments often drive not only the content of this program, as you’ve just witnessed, but also can help shape the content on Undecided with mothership. If you’d like to support us, please do subscribe. And don’t forget, you can Click the join button on YouTube.

That allows you to throw some coins at our heads. There’s also stilltbd. fm. There’s a become a supporter button there, and once again, it allows you to throw coins at our heads that lead to, well, usually just bruises, but sometimes something worse. Sometimes a minor concussion can occur like the one we just received from Wayne Palmiter.

Thank you so much, Wayne, for your generous donation. Thanks everybody for taking the time to watch or listen, and we’ll talk to you next time