¶¼ÁéÌåÓýÖ±²¥

Looking For More?

• Watch this episode on

• Connect with the Manufacturing Industry here www.amtonline.org/events

• Discover the past, present, and future of American manufacturing with

• Explore, watch, read, learn, join, and connect at

• Tune in to the AM Radio podcast

• For the latest in Manufacturing ¶¼ÁéÌåÓýÖ±²¥ news www.amtonline.org/resources

Ramia Lloyd:

Welcome to the Tech Trends Podcast where we discuss the latest manufacturing, technology, research, and news. Today's episode is sponsored by IMTS. I'm Ramia Lloyd and I'm here with...

Elissa Davis:

I'm Elissa Davis.

Stephen LaMarca:

AMT's roboticist, Stephen LaMarca, senior roboticist, lead roboticist.

Benjamin Moses:

The man behind the curtains, Benjamin Moses. See, let's get into that a little bit, but before we get into that, since we're talking about IMTS, let's talk about Chicago.

Ramia Lloyd:

Yes.

Benjamin Moses:

And we're going to be there for a little bit. I know Ramia was just there.

Ramia Lloyd:

Yes. So we went, it was a couple weeks ago, and we recorded our new series called Chicago Local. It's now an IMTS Plus. Go check it out. We did a couple of restaurants and we did a couple of hotels and then we did some neighborhoods. We walked around and we just talked about the history of Chicago and how all the fun things you could do when after you see all of our wonderful things at IMTS, what you should do after, and it was so much fun. First of all, I'm now a Chicago local, so me and Peter are neighbors. I'm the guy that he knows. No.

Stephen LaMarca:

He's got a guy.

Ramia Lloyd:

He's got a guy and now I have a guy. It was really fun. We got to see some of the really... I've never been to any of these restaurants, but they all were so good. We got to eat some of the food afterwards and I was just like... Michael Mark's doing all the work and he's filming, and I'm just carrying the stuff around his lackey. I was just secretly eating everything at the end. At the day so full. Everyone was like, "What's for dinner?" And I was like, "Dinner?"

Benjamin Moses:

I've been eating all day.

Ramia Lloyd:

[inaudible 00:01:38] been eating all day? Oh my God. It was so much fun. We started with Guinness. The Guinness New Factory is one of the two in the world-

Stephen LaMarca:

A US based factory.

Benjamin Moses:

[inaudible 00:01:49].

Ramia Lloyd:

Yes, in Chicago's. It's two. So there's one in Baltimore and there's one in Chicago, but the Chicago one is beautiful. They also have a bakery. So cool. The food is delicious. And then we went, where did we go after that? We went to Avec, so Donnie Madia. Donnie, yeah, Donnie Madia. Really cool person. He was in The Bear, really good food. He used to be the owner of Blackbird. It used to be a restaurant in Chicago.

Elissa Davis:

It's closer.

Ramia Lloyd:

So now he has Avec. Delicious. There's like a focaccia-

Stephen LaMarca:

Isn't it like a Michelin star restaurant?

Ramia Lloyd:

I don't know. Let me get back to you.

Elissa Davis:

I think Blackbird was.

Ramia Lloyd:

[inaudible 00:02:25] while you guys are doing your articles later.

Stephen LaMarca:

I'm sorry, I don't mean to be bougie.

Ramia Lloyd:

It's okay. I'm going to Google it while you doing articles later and I'll tell you.

Stephen LaMarca:

Avec means with in French.

Elissa Davis:

Yes.

Ramia Lloyd:

Oh, that's really good.

Elissa Davis:

I don't know a lot about French. I took three years, didn't pick up a single thing, but I know Avec means with.

Ramia Lloyd:

And with that, it was delicious.

Benjamin Moses:

Nice.

Ramia Lloyd:

Thank you. What did we do after that? And then we went to go to Monteverde, which is an Italian restaurant. It was so good. There was this homemade, it was like a biscuit and then she put prosciutto in it, and goat's cheese. Oh my God, it was so good, in a little sandwich. It was delicious. Oh my God. 10 out of 10 experience. And then we went to a bunch of hotels, which are all beautiful and you should all check them out when you go.

Benjamin Moses:

Nice.

Elissa Davis:

Well, everyone's finding out that Chicago's got a great food scene, because of The Bear.

Stephen LaMarca:

The Bear. I finally caved and got Hulu and I've been watching The Bear. And as if I'm not already excited enough to go to Chicago for IMTS, of course, I'm really excited for the food vibe again. I think The Bear does a really good job of meeting the bougieness of Michelin star dining with the rough, rugged, gritty, blue-collar vibe that is the people of Chicago.

Ramia Lloyd:

Yes. Absolutely.

Stephen LaMarca:

And this goes back to my first IMTS in 2016 when I fell in love with Chicago, because I'm a big fan of New York, both sports and just like the city. There's a lot of charm to be found underneath all that filth.

Elissa Davis:

I mean, Chicago is like New York's gruff uncle.

Stephen LaMarca:

But it's so much cleaner. And it was like, this is amazing. But I mean 2016 was a totally different time then, and one of my complaints was like, I love Chicago, it's just a little too clean.

Elissa Davis:

I feel like there's a lot of people who would say the exact opposite. So it's really interesting.

Stephen LaMarca:

Because they've never been to New York.

Elissa Davis:

That's true. New York in the summertime smells like hot garbage.

Stephen LaMarca:

Also, my favorite cities, like both pre working at AMT, my favorite cities were New York and Baltimore. And Baltimore's underrated. The Wire did it dirty. Let's remember that it's called the Charm City. Chicago.

Elissa Davis:

Chicago's it. Okay, so we're going to take a group field trip to Mr. Beef, which is what it's based on. Mr. Beef is the actual restaurant. They film the pilot there.

Stephen LaMarca:

Okay, cool.

Elissa Davis:

And it's what the restaurant is based on, the original restaurant. So we should take a group field trip there. I don't know where it is in Chicago.

Benjamin Moses:

We'll figure it out.

Ramia Lloyd:

It's not far from McCormick.

Stephen LaMarca:

Also, just a quick tangent. Carmie, the main character in The Bear has his resume, his big line item on the resume, is retaining a Michelin star while working at the French Laundry. I've heard of the French Laundry. I've never really looked into it. I've finally watched a YouTube video of some guy, I think his channel's, called The Guest. And he just goes to fancy restaurants and this guy has a different jacket in literally every shot.

Ramia Lloyd:

That sounds fun. I love that.

Stephen LaMarca:

This is leading where it's going. I watched this video and find out, fast-forward, I did see that the bill that for this two people, he didn't pay for his $1,600. But there was a lot of alcohol and a lot of old alcohol involved. It's like the wine alone is worth it when you look at that price. But this is just the bougie side of me speaking, so I'll shut up.

Ramia Lloyd:

That's my rent.

Stephen LaMarca:

But I was looking at the [inaudible 00:05:59], and just watching this video called me poor on so many different level.

Ramia Lloyd:

The video called you poor.

Elissa Davis:

To be fair, The French Laundry has a two year in advanced reservation requirement. It's insane.

Stephen LaMarca:

Yeah. And you look at the people that founded it, they're so humble and cute. I know their clientele is super bougie and those are the people you want to avoid, but the people that work there seem so incredible. And I know that's not Chicago and we're off-topic, but I just wanted to say that.

Elissa Davis:

Well, my whole thing about Chicago I've learned a lot is the improv, the theater scene in Chicago, which has also exploded in recent. I mean ever since Hamilton put one of their home stages in Chicago, I feel like it's definitely brought some more attention to the Chicago theater scene, which is really exciting. But Second City is one of my favorite places in the whole world.

Stephen LaMarca:

What is that?

Elissa Davis:

So Second City is basically one of the birthplaces of improv.

Stephen LaMarca:

Is it a theater?

Elissa Davis:

Yes.

Stephen LaMarca:

Okay.

Elissa Davis:

So they have multiple theaters within the building. Right?

Stephen LaMarca:

Okay, cool.

Elissa Davis:

But this is where a lot of SNL greats trained, but they also have a home base in Toronto. That's where Lorne Michaels knows a lot of the original SNL cast from, is from the Second City in Toronto. So basically there's three major improv groups in the United States. There's Second City, there's the Upright Citizens Brigade, which was founded by Amy Poehler and another fellow Second City alum, and then the Groundlings, which is in LA. So Second City is where you find the best improv, in my opinion. I mean obviously the Groundlings and Upright Citizens Brigade is great, but Chicago also has multiple other improv places. There's the Annoyance, which is another bar, and they do what's called skin-prov, which is basically people strip down to their underwear over the course. But it's like people who are new to improv to make them comfortable on stage.

Ramia Lloyd:

Do they have an audience?

Elissa Davis:

No, no, no. No, no, no. The improvisers.

Benjamin Moses:

The one on stage.

Elissa Davis:

People on stage, they're like stripping down.

Stephen LaMarca:

How is that going to help them?

Ramia Lloyd:

I feel like that would have the opposite effect, because now I'm naked and afraid.

Stephen LaMarca:

That's going the wrong direction, right?

Benjamin Moses:

It's a different sale.

Elissa Davis:

Yeah, my bad.

Stephen LaMarca:

From everything Nickelodeon taught me as a child, you're supposed to picture the audience being naked.

Elissa Davis:

I mean I saw it and it was better than I was expecting. There's a couple other ones.

Stephen LaMarca:

I believe it's amazing.

Elissa Davis:

And I can't remember their names right now, but The Annoyance and Second City are two of the big ones. So I'm definitely going to plan a group trip to Second City this year. I know there's more people who want to go this year.

Ramia Lloyd:

I'm upset that I missed it last year due to unforeseen circumstances.

Elissa Davis:

Due to your Uber kicking you out.

Stephen LaMarca:

That sounds like a different topic.

Ramia Lloyd:

I mean, you were there.

Stephen LaMarca:

It was a haze.

Ramia Lloyd:

Anyway, I'm in. If you do a group chat.

Elissa Davis:

Yes. Yes, we're definitely going to go.

Benjamin Moses:

Awesome, thanks. I'm looking forward to that. Yes. Speaking of which, can you tell us about today's sponsor?

Ramia Lloyd:

IMTS 2024, inspiring the extraordinary. From around the globe. The industry gathers to discover the latest in innovations of technologies changing our future through advanced and traditional manufacturing, robotics, automation, and digital transformation. Be a part of the change, register at imts.com.

Benjamin Moses:

Thanks Ramia. Steve, progress is being made on the test bed.

Stephen LaMarca:

Okay, so as lead roboticist at AMT, let's go back to academia, when you were in school. When you're throwing yourself at a really difficult topic subject or whatever, and it just doesn't make sense at all and then something finally clicks and you have that aha moment.

I had that again at the ripe age of 35 and it was exhilarating. I've gotten to a point, I've been able to control the robot and move it semi manually through the computer, do something on the computer and the robot moves, but I'm actively doing something on the computer. Finally got to a stage, and took long enough. Finally got to a stage where entered some things, and then hit the play button, hit go, and I actually got movement out of the robot. And that was the first major step, and it wasn't where I wanted it to go, and we've since had a few crashes, but that's not the point. Then this morning, I tested a program that I wrote last night and just made it very simple. I'm acting like this is lines of code. It might be like at least three lines of code. So we're not talking about, but I got the robot to move.

Elissa Davis:

Nice.

Stephen LaMarca:

Without running into any singularities. And I'm confident in being able to hit a target this week of being able to open and close the pocket and see enclosure. It's going to take a lot of effort, and it's going to just take a lot of time in throwing myself at it, but now I have 100% confidence, which is rare. 100% confidence in what I'm doing. And it's cloud nine right now.

Elissa Davis:

That's so exciting.

Benjamin Moses:

I want to hit on something that, because one of the things that we're running into is the arm getting to a specific pose and it basically locking up. It's hitting a singularity where it doesn't know what to do.

Stephen LaMarca:

And we just figured out the solution this morning, baby.

Benjamin Moses:

Nice. So I think that's a big takeaway that how we architected the layout, it's fairly compact because it's a factory on a table. And I think that's one of the drawbacks that we did in terms of planning, is we did a lot of physical planning. So we said the pocket [inaudible 00:11:49] is over here, the robotic arm has to be relatively close, but we never really got a chance to simulate all that before to see if we're going to run into problems. So, we're kind of doing stuff retroactively. So I think it's a-

Stephen LaMarca:

To basically do a quick speed run on what happened between the last episode and what happened now on the test bed, we were at a point where we were running into singularities a lot, but there's basically three ways to program the robot motion. You can either program the motion with respect to the base of the robot. So we're talking in Cartesian coordinates, or just a coordinate system in general. It's a mix of Cartesian, and polar, and or rotary coordinates. I forget what those are called. That's going to bother me. But with respect to the base, with respect to the tool at the other end of the robot, and with respect to the joints, programming a robot with respect to its base and our tool, yield you much smoother, much more fluid movements, and can do it in less lines of code.

Benjamin Moses:

That's cool.

Stephen LaMarca:

But requires higher, more complex math per line of code. And one of the things that Chloe, God bless her, has been throwing herself at is something that Nicholas, our German friend at IGIS, has been really trying to teach us and he's been doing a great job. We wouldn't have had this aha moment without him, but he's been trying to get us to do a circular motion with respect to the head or the tool of the robot. And that's been causing us to run into a lot of singularities, but it would theoretically yield a really smooth fluid movement. And part of that programming is, okay, you need to define your starting position, the closed position of the enclosure that the robot is going to grab onto, and not the open position because that's your target position. That's what you're trying to find, even though you know where it is.

You need to find the 180 degree position, which is the opposite side of the circular path that you're going of where the 180 degree opposite position of the closed position. That needs to be found and defined, even though it's out of reach and requires nothing but math to find that on your end. And assuming you do all of that right and enter in the coordinates properly into the robot, only then will you get a smooth cohesive movement. But we've also found out when the joints go from a positive degree and have to pass the zero degree movement to a negative degree, that's when the other joints are like, "Oh, this joint is approaching zero, let's twist it around to keep the speed." And then boom, singularity. It's an internal crash, basically, on the robot. And it's just the big robot companies, UR, FANUC, KUKA, they all sell software for their robots to avoid singularities.

And with the IGIS, I figured out, dude, we don't want to do that. We don't want to do these tool movement, or tool-based, base-based movements. When we're about to run into a singularity, we can avoid it by stopping that movement before a particular joint hits, is zeroed out, and then do strictly just a joint movement, just programming the individual joints alone, which is easier than it sounds, to get past that point of singularity. And then continuing a smooth, either linear, which is having the robot move either, probably the tool, in a straight line or a circular motion, about an axis that again has to be user-defined, not fun, requires math and measurements and just a lot of writing and whiteboard use, whiteboard real estate, or just controlling each individual joint, which is surprisingly easy. And that's what we learned today.

Benjamin Moses:

So the big takeaway, Steve, is to break the motion up into several segments.

Stephen LaMarca:

Yeah.

Benjamin Moses:

Okay.

Stephen LaMarca:

Yeah.

Benjamin Moses:

I just want to summarize it to make sure I understand.

Stephen LaMarca:

Listen, it took so much effort to get there. If I had just said that, you'd be like, "Oh, big whoop." It sounds like a three-year-old could do it. You're probably right.

Benjamin Moses:

But I'm summarizing in hindsight. That's all.

Stephen LaMarca:

I just want to shout out to Brooke Mewton at Kawasaki for being a huge inspiration. And Jessica at FANUC, I'm sorry I forget your last name. I'll reach out to you on LinkedIn. She was a huge inspiration. These roboticists have been really helpful. Nicholas of course, who actually helped with the software. And I also want to shout out to Kyle Salibi of Georgia Tech, which we all love. And Tom, Tommy Feldhausen, my boy over at Oak Ridge National Labs, both of them gave me... I'm not going to say they gave me crap, but they gave me little statics. Like, "Steve, you think programming a CNC machine is difficult? Dude, wait until you get in the weeds on programming a robot."

Benjamin Moses:

Yep. Now we are.

Stephen LaMarca:

Well, I'm there and I have successfully written a robot program and still to this day haven't written a successful CNC program, so it's officially easier. You are wrong. I'm smarter than some PhDs. No, I'm not.

Benjamin Moses:

We should move on pretty quick from there. Elissa, I know we've been talking about AI, which is going to be a big topic in general, but tell us what you found on, if it ever makes money.

Elissa Davis:

Yeah, so I mean that's a big question. Is AI ever going to make money? So this is a CNN business article and the title is "Has the AI bubble burst?" Wall Street wonders if artificial intelligence will ever make money? And this is because, so in the 18 months since ChatGPT kicked off the AI arms race as they call it in the article-

Stephen LaMarca:

I mean, that's pretty accurate.

Ramia Lloyd:

That's a solid title.

Elissa Davis:

Tech Giants have promised that the technology is poised to revolutionize every industry and it used its justification for spending tens of billions of dollars on data centers and semiconductors needed to run large AI models. Compared to that vision, the products they've rolled out so far feel somewhat trivial. Chat bots with no clear path to monetization, cost saving measures like AI coding and customer service, and AI enabled search that sometimes makes things up. And Big Tech is relatively slow to show for all their billions spent in terms of significant revenue. So really what it's saying is that these big companies like Amazon, and Microsoft, and things like that, it's like where is the big money that basically they promised from these AI creations? I know it's only been 18 months.

Stephen LaMarca:

Fair. I have thoughts. I have thoughts.

Elissa Davis:

Okay.

Stephen LaMarca:

Two things. First, I know I've been pro-AI, but I've been saying that the big risk in AI is not Terminator Two, it's the fact that these big money corporations are throwing themselves at racing to release the big first AI to market. And that's dangerous, because they are throwing money at the development and not throwing money at the safety of AI. That's a major problem. And hopefully the government does more to regulate that.And to some degree I know that there are a lot of good actors in the AI industry to making sure that that doesn't happen. But I do want to highlight, as somebody and seemingly a fanboy of AI, that is a major risk, and that is the risk.

Secondly, the fact that we're concerned about, well, AI is not making money yet. And it's like AI is a tool. That's like corporations crying about they're looking at their most advanced wrenches and be like, "Make money, make money." And it's like, no, you have to pick it up and turn it. You have to use it. It doesn't replace jobs. I say this all the time, because I feel like it's my quote that I've coined for myself, but nobody lost their job to a wrench. You lose your job to somebody that turns a wrench better than you. AI is just a tool. AI automation, digital tool, they're all just tools. Learn how to use the tools.

Elissa Davis:

Well, and that's what they're saying is that, so they've put these billions of dollars into these data centers and semiconductors and all they've got to show for it are these chat bots.

Stephen LaMarca:

So dumb.

Benjamin Moses:

You do hit on two key elements though I think we should definitely get into, or actually a couple elements. One, is the hidden cost of artificial intelligence and machine learning. The computing power required is astronomical. I mean, we still have cutting edge ships that are still, I would say, behind the times in terms of how fast they can process this stuff. And to your point, they're setting on massive data centers and the computing power required to do that, just to your point of a chat bot, is astronomical. And I think it is hitting on two different kind of business models.

One is we'll call big AI, large corporations racing to get what they consider a product and the revenue generated from the product, but what I would call local AI or local problems, I feel like that is where we're getting a lot of value. So for example, if I'm trying to solve something within my four walls using machine learning or advanced mathematics to solve that specific problem, get revenue off that and move on to the next problem, as opposed to this big ever-present AI model. I think those are two different business models.

Stephen LaMarca:

And not even solving the problem outright, but speeding up the process to a solution.

Benjamin Moses:

Yeah.

Stephen LaMarca:

From a writing standpoint, writer's block is the bane of any writer's existence. And a chat bot, an AI-powered chat bot might not, you're not going to cut and paste the paragraph that the chat bot sends out, but it will inspire you just a little bit to help you overcome that hurdle that is stopping you from continuing and will be like, "No, rewrite it like this dummy." And then you've got it, you're past it. It speeds up the whole like, oh my god, sitting on the couch watching YouTube, or playing games with friends, because I don't know how to continue what I'm trying to say. Instead, you just throw yourself at a chat bot. You have a few iterations with chat GPT, or perplexity, which shows you sources. And then you're through it.

Benjamin Moses:

I do want to hit on that last point, but-

Stephen LaMarca:

It's a tool.

Benjamin Moses:

One of the big hurdles I think for most ever-present AI tools is the trust. I still think there's a lack of understanding of how did you get there? For you it's some basic writing prompts to guide your writing. You can kind of see the iterations, you can see what you posted versus what you asked for, but if you're asking larger questions where you just get a result, why would you ever trust that from anyone?

Stephen LaMarca:

It's like when you're in academics, whether it's math, or history, or something, if you're producing a deliverable to your professor or teacher, they don't just want to see the problem answered. They don't want to just see the essay. They want you to show your work, they want you to cite your sources. And then AI platforms like Perplexity, they're like hard-lined programmed to, these are the first things you provide, the sources and you show your work, then you explain your answer.

Elissa Davis:

I do want to say real quick that the Microsoft CFO Amy Hood said that we probably won't see monetization of AI technology until 15 years or more.

Benjamin Moses:

Oh, man.

Elissa Davis:

And the Meta CFO, Susan Lee basically said the same thing.

Benjamin Moses:

Wow, that's wild.

Elissa Davis:

We're looking at a long term investment of 15, 20 years and that's why investors are upset. Because they're like, "When we invest in these things and these public companies we're expecting this fast return."

Stephen LaMarca:

But who is seeing the fast return that nobody's talking about? The individuals using these AI tools. I'm holding my job, because I have found, I'm sorry, found how useful AI has been. It's helped me be a better worker.

Elissa Davis:

Yeah, but it's not making the company's money.

Stephen LaMarca:

It's making me money. It's helping me retain making money.

Benjamin Moses:

So I definitely want to move on to humanoid robots. We've been talking about this briefly and I'm still skeptical about it, but BMW tests humanoid assembly robots from a company called Figure AI. This is from Design News. And on the surface I'm like, "Fine." We've been talking about a lot of companies kind of testing that. I think they're trying to figure out is this a thing to latch on? Kind of the AI tool is just the thing that they want to do. And there's a couple of key elements actually that BMW hits on, where there's really fascinating lessons learned about the iterative process towards humanoid robots. I'm still skeptical, but they definitely hit on the integration of the communication of the robot itself, back to the overall system. So they are actually solving-

Stephen LaMarca:

Data.

Benjamin Moses:

... Autonomous problems, but you're exactly right. They're running into what data needs to be communicated, and how to communicate that wirelessly.

Stephen LaMarca:

Any new robot being brought to market is going to be useless unless the user and owner of it can track the digital thread, the data stream coming off of that robot, to make sure it's doing what it's supposed to be doing.

Benjamin Moses:

And it's another segment that they hit on is the processing power with the additional data that they want to start processing. So not only are they looking at, of course, similar to a co-bot where they have force feedback, they have vision systems, they've got microphones, they've got increased cameras, they've got communication capability. So all of that has got to go through the robot itself and go back to a server or some edge computing to process. So, the bandwidth and the computing power is growing exponentially in some of these factories. And the last thing they talk about are what are the capabilities, what can a robot do? So I think it's a really interesting dive from BMW to like, okay, this is a cool thing, what are the problems we're going to run into and what are the capabilities?

Stephen LaMarca:

I think this is really interesting and I want to add on to this, because a couple months ago, it feels like a couple weeks ago, but a couple months earlier this year, if not late last year, we were talking about, well, the CTO at Boston Dynamics said that humanoid robots aren't going to solve everything. Them being somebody that also makes humanoid robotics, but they've also gone to other legged robots. And when Amazon was experimenting with agility, robots Digit, their humanoid robot. They were like, "For warehouse work?" Especially, if all of the warehouses are single level and they're not going up and down stairs, why would you use a humanoid when you could just get a conventional IMR, an industrial mobile robot, an AMR, an AGV with a robot arm on top of it and pushing around a bin? That's what you need. You don't need to get fancy with a humanoid.

But what's interesting about that is contrary to what that Boston Dynamics chief executive said, and there's a lot more companies coming out as of late releasing humanoid robots. So maybe we're not seeing it, but there's a demand for humanoid robots there, because Tesla has released their humanoid robot. I forget the name of it. And now Nvidia has come to market with a humanoid robot. The more important news there is Nvidia has released a new silicon chip specifically for controlling humanoid robots. So it's like, what do they call them on motorcycles? An IMU, an inertial management unit, six axis IMU? They have those specifically for robots now, and I wouldn't be surprised if that silicon finds its way to the motorcycle industry.

Elissa Davis:

So it kind of sounds like similar to the AI thing is that it's a race to see who can get the best humanoid robot the fastest.

Stephen LaMarca:

Right. Well, at least-

Elissa Davis:

Because why does every company need a humanoid robot?

Stephen LaMarca:

At least NVIDIA's intentions are making money selling superheated sand silicon. At least we know what they're up to.

Benjamin Moses:

It's a scenario where they have a technology that they're trying to find some of the problems and that they can solve, but also solve some of the inherent problems with early development of, so let's not be saying human robots is still an early development, it's not a mature product.

Stephen LaMarca:

No, they're not far along.

Benjamin Moses:

So they still-

Elissa Davis:

The news makes it seem much farther along than it is.

Stephen LaMarca:

Right, because everybody, all of the 24-hour news cycle, the fear-mongering, wants to see Terminator happen.

Benjamin Moses:

So they still have a ways to go to get past the hype cycle, the trough disillusionment to weed everyone out, get past the fast money. So, there's a lot of that where they're trying to get early investment and fast money out of this thing that we'll probably solve problems. We're not sure what problems are going to solve just yet. And it is scenario where you have to look at the different segments where they're applying it. So at a massive factory, sure, maybe, but then how does that trickle down to a contract manufacturer that's like 20 million in revenue, 10 million in revenue. Those are a lot of different scenarios where maybe the humanoid robot doesn't necessarily apply, but maybe technologies from that robot may apply.

Stephen LaMarca:

Yeah.

Elissa Davis:

That's fair.

Stephen LaMarca:

Elisa. I think we should talk about, are AI's conscious?

Elissa Davis:

So this is from an article from Psychology Today, which I feel like there'll be some scientists who probably already know, like mathematicians who watch this or listen to this podcast will be like, "Psychology Today?" But I'm a psychology nerd, so I felt as I saw this and I was like, "That's really interesting." But basically it boils down to... So hold on, let me find the actual, okay. "A living organism is a never the same physical and psychological entity from one instant to the next. The computer hardware on the other hand remains the same physical structure from the moment it leaves the factory until it stops working or is discarded."

Benjamin Moses:

Interesting.

Elissa Davis:

So by that principle, AI will never reach a human level of consciousness.

Benjamin Moses:

Because it stays stagnant.

Elissa Davis:

Yes, exactly. Because it can become smarter, right? And it can become more efficient and it can do all these things, and potentially it could learn emotion, but it will never actually be human, so therefore it can never actually be conscious, if that makes sense. I'm sure we still run into the fear of Ultron, that kind of thing happening.

Benjamin Moses:

I think it's a very relevant conversation because there's a relationship back to, the term is called hallucinations, within a lot of the chat bots. So when you ask it a prompt, it starts varying in a different direction where it's completely unpredictable. And I think there's a kind of correlation between the trust model, and one, the basic understanding of this massive AI tool. Do we ever understand how this works? And to your point of the hardware may stay stagnant, but what is the definition of iterations or change of an AI model and does that parallel human thinking, and our ability to rationalize? Does the AI model exist as a consciousness or is it stagnate code?

Elissa Davis:

Yeah, exactly. It can only think as far as the code can go.

Benjamin Moses:

Right.

Elissa Davis:

Right?

Stephen LaMarca:

This is interesting because it makes me think of that time Google fired somebody for having claimed that Google's AI is now, what's the word? Sentient.

Benjamin Moses:

Right.

Elissa Davis:

Yeah. But again, it's like-

Stephen LaMarca:

That AI developer says differently from this article title.

Elissa Davis:

But I think what we have to consider is the subject nature of consciousness of sentience. Do they think it's sentient because it's saying that it can feel things, or is it saying that it's sentient because it can actually feel those things? So because it doesn't experience human life, it doesn't experience human nature or human experiences, it will never fully reach a consciousness that we have as humans.

Ramia Lloyd:

So the first AI conversation made me think of this thing that I saw on the internet, and have you guys seen, it's a necklace that you wear and it's an AI in it, and it's called The Friend.

Elissa Davis:

No.

Ramia Lloyd:

In the commercial that I watched, they were talking to it and they would text them, and they would be watching something, and then the AI necklace would text them and be like, "Isn't this such a funny show?" I don't know. But as soon as you started talking about it, I was like, "Holy crap." Have you guys ever seen the movie Her, where he falls in love with his AI, who's like Scarlett Johansson? Which is understandable, but it's so confusing and it's so creepy. And it made me think about it, because you just said how they never experienced full human emotions, and its getting, I am kind of afraid that it's going to get there. It's kind of scary.

Stephen LaMarca:

It is.

Elissa Davis:

I can understand why people think that. I think for me, I don't know. Because again, it goes back to code. It all boils down to this code. So if it's not coded to feel those things, it's never going to feel those things.

Stephen LaMarca:

Right. You and your article make a fascinating point, because I feel like I don't belong here. We should bring in an IT, like Sean from IT, who is a network admin. Be like, "Well, you're only going to get a signal from another computer on the network if you ping that computer first, it's not going to actively ping you on its own."

Ramia Lloyd:

Yeah.

Stephen LaMarca:

[inaudible 00:33:57] people.

Benjamin Moses:

I think one of the interesting thing about AI is that the concept itself is never the same. What we considered AI, and even for video games, the bot that I'm shooting at, I still call an AI, right? It's not the same type of AI that you're going to ask ChatGPT or Gemini or anything like that. But our perception of what we consider AI is fascinating, and it's involved significantly over most of our lifetime. And I think there our concept of consciousness related to AI I think may evolve a little bit too. But I don't know, it's just interesting to see this term itself doesn't have a firm definition, because it constantly changes. And our perception of AI seems fairly stagnant also.

Stephen LaMarca:

Computers are better, computer programming is better. The general public cannot wrap their head around why they are better. We call it AI.

Benjamin Moses:

Right.

Elissa Davis:

That's fair. Yeah. Personally, I don't think AI is going to ever going to be able to take over the world, just from experiences I've had with AI where I tried to talk to a chat bot and I'm like, "You're not understanding what I'm asking you." I have a cash bond from when I was a kid. My mom gave it to me, because after 30 years it stops accumulating money. So I sent it into the Treasury, but originally I was like, "Oh, I can do it through my bank." And so I went into the USA app and I was typing in, "Cashing bond" and it kept trying to send me to investment stuff. And I'm like, "No." And I tried four different times and I was like, "This AI does not understand what I'm asking it." And again, that comes back to coding and I know that, but it's like, I don't know, because what humans feel and what humans experience is constantly changing, we can program that into an AI, but then that will be outdated before we even realize it.

Stephen LaMarca:

Yep. Maybe the general public is just nervous about the day that they can't use the nuclear option, which is hitting zero when talking to a phone bot. They're like, "I want to talk to a human." It's like, "Well, we need to know which department you want to talk to?" "Zero, zero, operator. Let me talk to a human." And then one of these days they're going to send you to a voice activated LLM. And then what's going to be the zero for that? How do you hit zero when talking that thing?

Benjamin Moses:

We'll find out.

Ramia Lloyd:

I don't know, but I'm going to be nice to all of them.

Stephen LaMarca:

Be nice. Absolutely. No throwing your phones After you query check GPT.

Ramia Lloyd:

I always say please and thank you.

Stephen LaMarca:

Thank you so much. You've helped me so much tonight.

Ramia Lloyd:

I love you.

Stephen LaMarca:

Awesome, guys, thank you. Let me know if you have any more questions.

Benjamin Moses:

I think a lot of the topics we hit on are really good feeders for preparing for IMTS, so I'm definitely excited for upcoming show.

Elissa Davis:

It's going to be a great show. And we'll be live at IMTS, right?

Benjamin Moses:

We will be.

Elissa Davis:

Yeah.

Ramia Lloyd:

A couple of times.

Stephen LaMarca:

We've got more episodes before IMTS, right?

Benjamin Moses:

We'll talk about it.

Stephen LaMarca:

Okay.

Ramia Lloyd:

Ooh, the suspense.

Stephen LaMarca:

We didn't touch you on open USD. And on that note, it's time to log off.

Benjamin Moses:

Ramia where can they find more info about us?

Ramia Lloyd:

Amtonline.org/resources like share, subscribe.

Benjamin Moses:

Bye everyone.