Looking For More?
Watch this episode on YouTube
Connect with the Manufacturing Industry here www.amtonline.org/events
Discover the past, present, and future of American manufacturing with www.mmsonline.com/madeintheusapodcast
Explore, watch, read, learn, join, and connect at www.imts.com/
Tune in to the AM Radio podcast www.additivemanufacturing.media/zc/am-rad…o-podcast
For the latest in Manufacturing Technology news www.amtonline.org/resources
Transcript:
Ramia Lloyd:
Welcome to the TechTrends podcast, where we discuss the latest manufacturing technology research and news. Today's episode is sponsored by MFG Conference. I'm Ramia Lloyd, and I'm here with ...
Elissa Davis:
I'm Elissa Davis.
Stephen LaMarca:
I'm Stephen LaMarca-
Bejamin Moses:
And, Benjamin Moses. What's up everyone? Welcome to 2025.
Elissa Davis:
We did it.
Bejamin Moses:
Merry New Year.
Ramia Lloyd:
How exciting.
Bejamin Moses:
I want to get right into some technology stuff we've been working on, because we made some really good progress over the break, or my break, because I took a long time off.
Stephen LaMarca:
Listen, I didn't do anything. It was all on Chloe this time around, because she genuinely enjoys working on the test bed, and who doesn't? What's not to love? Chloe has been in charge of sourcing and implementing the vision system that will be employed on our test bed, and another thing that was done over the break that she did was installing Nvidia's Omniverse and getting it up and running, and that piqued my interest because the most recent thing that I did that was really big on the test bed was sourcing and assembling, with a lot of help, the new PC. I want to make sure that it's performing to spec, so whenever she runs or anybody runs Omniverse, I was like, "This is what we built this thing for, so I want to make sure it's performing," and we're doing benchmarking.
She's been doing a lot, mostly with the vision system, and she told me the other day that she threw herself, for eight to nine hours, at training the data set for this vision system, which is an Intel RealSense depth sensing camera, meaning there's two cameras that use parallax to determine how far something is away, just like your own eyes, but you have to take a lot of pictures, a lot, a lot of pictures, and for each and every single one of those pictures, you need to, with a cursor, select, make a little red box around the things that you want to identify. In our case, we are identifying different materials of bar stock, and we've got different types of six-inch long, one-inch diameter bar stock that can work in the Pocket NC of three different materials, Delrin, aluminum, silver ... Delrin being white plastic, aluminum being silver colored, dull metal, and brass, yellow gold-ish in color, metal.
For every, I think she took hundreds, if not thousands, of pictures, and for each and every single one of those pictures, she had to put a box around each of the pieces of bar stock and identify it as what type of material it was, and she did that for eight or nine hours. Another cool thing that she told me about doing, just to build the data set so it knows what it's looking at, so it can learn from that data set, another cool thing that she told me about is once it starts looking around, the camera starts working on its own and recognizing things on its own, one of the sliders, one of the parameters that you adjust in the programming, funny enough ... Well, at least I found it amusing. It's called confidence, and you adjust the confidence from zero to 100%. Immediately, on hearing about this, I was like, "Well, who doesn't want to be 100% confident all the time?" Problem is, that I did not realize, but it seems pretty obvious now, if you are at 100% obvious or 100% confidence, you are going to be 100% wrong all of the time, most of the time.
It's funny that she found that the sweet spot for accuracy for a camera with the setup that we have now, and some of the variables that affect this, affect the accuracy, are the lighting, and we do not have optimized lighting over there, kind of like we do in the studio, because it's right next to a big window. Window means daylight, and the day is constantly changing because time is not fixed. We don't have to get metaphysical, but the light is constantly changing, so it's constantly skewing the data. It is a variable that can be accounted for eventually, and I don't want to put that in her eyes right now or in her mind right now because she's got a million other things she's working on, but she found that the ideal confidence setting for to get the most accuracy out of the image recognition is four to 8%.
Bejamin Moses:
Confidence level.
Stephen LaMarca:
Four to 8% confident.
Bejamin Moses:
Yeah.
Elissa Davis:
Wow.
Stephen LaMarca:
I want to be more like this camera.
Bejamin Moses:
Well, it's interesting as you go through this, right? The process of building a training set for object recognition, it's laborious, but it is fairly straightforward. You're taking pictures of different scenarios and you kind of articulated the different factors, right? Lighting, changes of color, change of shadows, things like that, right, it changes what the image looks like. Training, it does take a lot of time, especially if you're working on something new and kind of unique for us, where you're identifying raw material. It's not a thing you can just download off of a data set that's commonly available, right, so creating custom data sets is a thing. The structure of light is fairly important, but the mathematics behind it, right, you talk about accuracy and confidence, which are two different things, right?
Stephen LaMarca:
Yeah.
Bejamin Moses:
That's a key. When you talk about the confidence, statistically, and you say, "Okay, four to 6%," everyone would be thrown off, but realistically, it's not bad, right, as a starting off point to get accurate predictions on what the camera can see. I think as we get more involved with AI, the importance of what the math actually means is going to be more relevant to us on the day-to-day basis. Right?
Stephen LaMarca:
Yeah. It's just, I found it really amazing. I think it's cool. It's cool learning a new thing like this, especially through somebody else that's doing all the hard work, but to quote that Michael Jordan, 30 for 30, or whatever video it's from, when Michael's, "I took that personally."
Elissa Davis:
"I took that personally."
Stephen LaMarca:
Because, with the confidence thing, I remember growing up, I heard from my sister a lot, "Oh, Stevie, often wrong, never in doubt."
Elissa Davis:
I would love to ... I mean, I also function at about four to 6% confidence on a daily basis.
Ramia Lloyd:
Same.
Stephen LaMarca:
Up it to eight. Try upping it to eight.
Elissa Davis:
I'll try upping it to eight, but it's definitely, I relate to the computer in that my confidence is also a sliding scale, but yeah.
Bejamin Moses:
Well, speaking of which-
Ramia Lloyd:
It also depends on the day.
Elissa Davis:
It also depends on the day.
Ramia Lloyd:
The way the sun's shifted.
Bejamin Moses:
Yeah, was going to say, yeah, exactly.
Elissa Davis:
It also depends on if the weather, just like the camera. Yeah, so basically I am this camera.
Ramia Lloyd:
I am. Same.
Bejamin Moses:
Speaking of which, I do want to hit on two things before we get into articles, is, it snowed, and personal goals for 2025.
Elissa Davis:
Yeah, it snowed.
Bejamin Moses:
I'm wondering, in our area, we're in the Midwest. No, not Midwest, middle Atlantic area.
Stephen LaMarca:
Mid-Atlantic.
Bejamin Moses:
It snowed. We got a fair amount of snow in our west of DC area, about up to six inches.
Stephen LaMarca:
In Fairfax, we got 10.
Bejamin Moses:
Shut down everything for the past few days.
Stephen LaMarca:
Yes.
Bejamin Moses:
I'm wondering if we ever get to a point where ... This is mid-inclement weather. It's not severe, it's not a typhoon, it's not a hurricane, it's not an earthquake. Six to 10 inches, a fair amount of snow, but it's not catastrophic. The government shut down. Opium was shut down, school was shut down.
Elissa Davis:
This all has to do with infrastructure.
Bejamin Moses:
Okay.
Elissa Davis:
Let me explain.Having lived in Colorado for six years, I lived in a place where it snowed a lot. It literally snows October through May, and they think they're prepared for snow because they have the snow plows out and they put sand down on the roads, which, mind you, sand doesn't really do what they think it does. It just makes your car really dirty. It doesn't really melt the ice.
Bejamin Moses:
Just makes it muddy.
Elissa Davis:
That's why there's a carwash on every block in Colorado.
Stephen LaMarca:
That scratches up your paint.
Elissa Davis:
Yeah, but I remember, so I went to high school, I graduated high school here in Fairfax County, and I remember one day it started snowing in the middle of the day, and it was big fat flakes, but it wasn't that bad, right, but they sent us home in the middle of the day.
Stephen LaMarca:
Fairfax?
Bejamin Moses:
Early dismissal, yeah.
Elissa Davis:
Yeah, and they got everyone on the buses and there was plows out. I was like, "There's nothing on the roads yet, even. Why?" I feel like in these situations we're either really over-prepared or really under-prepared.
Bejamin Moses:
Yeah. It's never a balance is it?
Elissa Davis:
Exactly, but I will say, in terms of school, so this is probably an urban legend, but my mom used work in Spotsylvania county, and she was a teacher there, and she said that part of the reason why schools here in Virginia, or at least from Fredericksburg up, are kind of cautious, is that one time there was a kid who ... They didn't cancel school even though there was ice on the roads, and some kid got hit by a car, and they could potentially be liable if that happens and they were going to school and that kind of stuff.
Bejamin Moses:
Yeah, way to the bus stop.
Elissa Davis:
Yeah, so that made them overly cautious, and so now at the slightest dusting of snow, they will cancel at the chance that there is ice on the roads.
Stephen LaMarca:
Fascinating.
Bejamin Moses:
That's fascinating because this area is notorious for bad drivers.
Ramia Lloyd:
Oh, God.
Bejamin Moses:
Yeah, so we have bad drivers.
Ramia Lloyd:
I just going to say, we have bad drivers on a normal day with an extreme amount of traffic. You add snow to it, and I'm glad. Everybody sit down. You're not going to be helpful to anyone in this situation. You already such at driving. It's just going to get worse.
Elissa Davis:
Also, people here, and I think it's because there's so many people from so many different places,
Stephen LaMarca:
So many different cultures from different parts of the world.
Elissa Davis:
Yeah, who drive here.
Stephen LaMarca:
Many different driving styles adhering to one law of driving.
Elissa Davis:
Yeah, and they don't know, a lot of people. There are people who don't know how to drive in the rain. There are people who don't know how to drive in the snow. There are people who don't know how to drive in the sunshine.
Ramia Lloyd:
Exactly.
Stephen LaMarca:
California does not know how to drive in the rain.
Elissa Davis:
No.
Stephen LaMarca:
People from Washington probably don't know how to drive in the sunshine. They're probably blinded. They probably don't have sunglasses.
Bejamin Moses:
"What is that?"
Ramia Lloyd:
"What is this?
Elissa Davis:
It is very overcast.
Stephen LaMarca:
There's people from parts of the world that are probably not used to having lines on the road.
Ramia Lloyd:
Yeah.
Elissa Davis:
Yeah.
Stephen LaMarca:
Lot of different cultures and styles.
Bejamin Moses:
Guys, 2025. What are you guys doing differently? Personal goals?
Ramia Lloyd:
My goal is to be 8% confident.
Stephen LaMarca:
8%?
Elissa Davis:
8% confident.
Bejamin Moses:
Up from two?
Ramia Lloyd:
From two to eight, we'll get there.
Elissa Davis:
Same.
Bejamin Moses:
I think a lot of mine just revolves around doing stuff I already did well last year, continuing, because I was off for basically three weeks. I didn't do anything good for my body.
Elissa Davis:
I mean, same.
Ramia Lloyd:
That's what vacations are for.
Elissa Davis:
Same, Ben.
Bejamin Moses:
I didn't read. I didn't do any exercise. I mean, it was great. It was a good time to be off. I stepped on the scale this morning. I was like, "Oof, 2025 is starting off heavy."
Elissa Davis:
The secret? Don't use a scale.
Bejamin Moses:
Just ignore it.
Elissa Davis:
Yeah.
Ramia Lloyd:
We went on a cruise for Christmas, and when we came back, my mom was like, "Oh, we should all go on a walk and we should do these things," and my dad was like, "We've just been on a boat for eight days eating and drinking, and now you want to do all this exercise? We could have done this on the boat if you were so serious about it."
Stephen LaMarca:
They have tracks to do that.
Ramia Lloyd:
We all just looked at each other and then we went to dinner.
Bejamin Moses:
I made all these plans. We spent the week in Williamsburg after Christmas. All these plans. I picked a place that has an indoor pool. We go swimming. It's got a good exercise area. We just lounged around. We barely even made it outside to do shopping.
Ramia Lloyd:
Yeah.
Elissa Davis:
I personally believe that vacations are not for exercise.
Ramia Lloyd:
Same.
Elissa Davis:
Unless you're going walking around in an amusement park.
Bejamin Moses:
We did that. I tried to do hiking, but-
Elissa Davis:
That's exercise. That's exercise.
Bejamin Moses:
Family was against that.
Elissa Davis:
You get your steps in.
Bejamin Moses:
I mean, I drank a lot of beer too.
Ramia Lloyd:
Oh my God.
Elissa Davis:
I mean, as long as you drink a lot of water too, it's fine.
Ramia Lloyd:
There's water in beer.
Elissa Davis:
Yeah, exactly. It's fine. My goal, I'm starting grad school.
Bejamin Moses:
Nice.
Elissa Davis:
My goal is to hopefully be good at it.
Stephen LaMarca:
I am sure you will be.
Ramia Lloyd:
Solid goal. I like it.
Stephen LaMarca:
Telling us about your class before we started recording, it sounds like you're-
Ramia Lloyd:
Yeah, it's right up your alley.
Stephen LaMarca:
You're going to do just fine, if not more. That's an understatement.
Elissa Davis:
Thank you.
Stephen LaMarca:
Going into 2025, I want to open my ears more, close my mouth more, try to keep my head down and vision clear for what I want to work on. Also, just not try to get overwhelmed, and not try to get too gung-ho, because 2024 was a rough year for a lot of people. I'm blessed that it wasn't terrible for me, but it went by really fast. I probably sound like a broken record from other years of our podcast because I say the same thing. I really want it to slow down and it ends up not happening. It ends up being the opposite of that. As we transitioned into 2025 during the holiday break, I saw a meme that really resonated with me hard on Instagram, and it was a picture of the summit of Mount Everest, and I think the quote was ... It's not verbatim, but the quote was something along the lines of, "Every dead body on Mount Everest was once a highly motivated individual. Maybe calm down." I'm like, "Dude, I love this." I don't think I'm the one who needs to hear it. I should probably get my steps in, but I really want more people to hear that, because we all at least know somebody that needs to see that meme.
Elissa Davis:
Yeah. Can I just say that that is the most millennial sentence I've ever heard? "I saw this meme that really, really resonated with me on Instagram."
Stephen LaMarca:
I am what I am.
Elissa Davis:
Yeah.
Bejamin Moses:
You're a walking meme?
Elissa Davis:
I mean, as a fellow millennial, I get it.
Stephen LaMarca:
When in Rome, be American.
Elissa Davis:
Sorry?
Bejamin Moses:
What's the problem? That's a good transition to our sponsor.
Elissa Davis:
That was really good. Okay.
Ramia Lloyd:
The MFG meeting brings together manufacturing technology industry leaders to address key business challenges and provide actionable solutions. Sessions cover a wide range of topics including workforce, economics, supply chains, leadership, politics, and more. The unique meeting experience provides unparalleled opportunities to network with executives from companies that make sales, service, and support MT. These peers and competitors will learn from each other, build long-lasting connections, and share insights that will benefit their business. Go to amtonline.org/events to register.
Bejamin Moses:
Awesome. Thanks, Ramia.
Ramia Lloyd:
Thank you.
Stephen LaMarca:
Hey, you know what else, something I also learned for 2025?
Bejamin Moses:
What's that?
Stephen LaMarca:
I actually learned this a while ago and I forgot to implement it earlier, but you should like, share, and subscribe, and comment. We should say that in the middle of the video instead of at the end. Apparently it gets better results in the middle.
Ramia Lloyd:
Oh.
Stephen LaMarca:
Do it.
Elissa Davis:
Do it. Yeah.
Ramia Lloyd:
Pause, everyone. Thank you.
Elissa Davis:
Like, share, and subscribe.
Stephen LaMarca:
Yeah, Ramia's dad. How about you leave us a comment?
Elissa Davis:
Yeah, tell us how you really feel, Ramia's dad.
Stephen LaMarca:
We hear you talking all this smack.
Elissa Davis:
Nice comments.
Stephen LaMarca:
Why don't you put it in the comments?
Ramia Lloyd:
Make them nice, please.
Bejamin Moses:
Let's get into some articles about da Vinci.
Elissa Davis:
Yeah.
Bejamin Moses:
He's still around.
Stephen LaMarca:
Ramia's dad.
Elissa Davis:
Yeah. Ramia's dad is actually Leonardo da Vinci.
Ramia Lloyd:
I wouldn't be surprised. He's kind of really smart and scary.
Stephen LaMarca:
All right, we're going to get him too full of himself.
Ramia Lloyd:
He's going to be at home texting me like, "Ha, ha, ha, ha."
Elissa Davis:
Now there's no reason for him to not leave a comment. Now he has to leave a comment, but-
Ramia Lloyd:
Comment if you're not da Vinci.
Elissa Davis:
Yeah, comment if you're not da Vinci. Okay, so I kind of want to preface this with a short story. Is that okay?
Bejamin Moses:
Sure.
Elissa Davis:
Okay, so in sixth grade I had a science teacher and he was also my homeroom teacher. He was obsessed with Leonardo da Vinci. We literally had an entire unit where we had to do a project about Leonardo da Vinci, something that he did, and we had to do a project about it. Ever since, I've been kind of interested in da Vinci, because I'm like, "Okay, well, if my science teacher's this obsessed with him, then he's probably pretty cool." Right? Obviously he's done more than just science stuff. He's also done art stuff, which is pretty cool, but also this is specifically about-
Ramia Lloyd:
He's also a ninja turtle.
Elissa Davis:
It's called da Vinci's bridge, and it was brought to life in the city of Bari, I think is how you pronounce it, B-A-R-I. They 3D printed it, so he designed it, and it's a self-supporting bridge. The arch is designed in a way in which it's self-supporting, so it doesn't need any suspension or anything like that. He originally designed it to go over the Golden Horn, which is an inlet of Bosphorus in Istanbul, Turkey. It was designed to connect the Para district with Constantinople. What they did is, this was the Polytechnic University of Bari and WASP, which-
Stephen LaMarca:
White Anglo-Saxon Protestant?
Elissa Davis:
No, it's something else.
Ramia Lloyd:
[inaudible 00:17:39], so confident.
Elissa Davis:
They actually used waste materials from stone processing to 3D print this bridge. They used a combination of low environmental impact mortar developed by the Italian startup B&Y, and that's made up of waste stone powders combined with a lime-based binder, providing a sustainable way to reuse the byproducts of marble and stone processing.
Stephen LaMarca:
That's cool.
Bejamin Moses:
Wow.
Ramia Lloyd:
That is really cool.
Stephen LaMarca:
All of the chips, scraps, and dust and chunks to come off of marble sculpting, the Carrara marble sculpting, which is what Italy's known for, that can all be ground down to equal-sized granules to be then 3D printed. That's so cool.
Elissa Davis:
Yeah, because basically they turn the waste materials into a concrete, right, and then print it that way.
Stephen LaMarca:
Will the 3D printed marble lose its marbling?
Elissa Davis:
Probably.
Bejamin Moses:
Does it lose its soul as it becomes a structure? Is that how it happens?
Stephen LaMarca:
Does the marble have good marbling?
Bejamin Moses:
Actually, you bring a really interesting point, where it's a good intersection of the concept of using waste material in 3D printing. That's been a growing concept, people printing at home, like, "I've got all this extra filament, what do I do with this?" Chopping it up and then reusing that chopped carbon fiber, it's basically that, right? You have all this waste material, let's reuse it into some type of material, in some type of form.
Stephen LaMarca:
I love watching the videos of people taking their disposable plastic bottles from buying water at Walmart. You take the label off, you clean up all the adhesive that was on the label, and then they 3D printed this little device that actually has an exacto knife on it or something like that, and it cuts up the plastic bottle, sorry, cuts up the plastic bottle to turn the plastic bottle into filament. That is so cool.
Ramia Lloyd:
Okay, wait. When I was a kid, we would put all of our plastics into a larger plastic bag, probably not safe, but we would reuse it, and then we would take it to the store and put it into machines that did that, and they would give you change and money from it. This was in Michigan, so when we moved to Tennessee, that wasn't a thing anymore, and we were like, "What do we do with all this plastic?" Apparently it's only, not very many states do this.
Stephen LaMarca:
No, yeah. No, Virginia doesn't do that.
Ramia Lloyd:
Doesn't make sense, because they would literally smash it and it comes into those cubes. It's crazy.
Stephen LaMarca:
It blew my mind. When I went to school in Vermont, there was this gas station that we lovingly referred to as Shady Mobile, near to the college. They had a deli, they had liquor-
Ramia Lloyd:
In the gas station?
Stephen LaMarca:
And you would bring them your bottles and cans and they would give you money, and then you'd spend it in the liquor store. Oh, you could also get gas there.
Ramia Lloyd:
Yeah, see? That's [inaudible 00:20:26].
Stephen LaMarca:
Wild. Vermont's efficient. Also, snow up to your waist, never canceled anything. Vermont doesn't cancel.
Bejamin Moses:
There's also five people in Vermont.
Stephen LaMarca:
There's also five people, and also, everybody-
Elissa Davis:
Just carpool everywhere.
Stephen LaMarca:
Everybody drives, all five of them drive a Subaru Outback.
Elissa Davis:
Oh, so they're fine.
Stephen LaMarca:
They're not getting stopped.
Elissa Davis:
No, they're fine.
Bejamin Moses:
Exactly.
Elissa Davis:
I do also want to mention, this is a pedestrian bridge. It's not a bridge bridge, so it's a small-scale model.
Bejamin Moses:
No, it's beautiful, but the idea, I think it was about five, six years ago, there was a Scandinavian company I was looking at, metallic 3D printing a bridge using robots. Right? There's a couple of concepts that I think bridges are a really good test case for universities to apply these applications, because one, it is useful to see. It's big, right? Can you span this structure? I mean, if they did a cup, that's not that interesting, right, but the ability for them to push their boundaries, both of the material and in the overall form, of a bridge, it seems like it's a repeating use case that seems to be testing the waters quite a bit. I really like this a lot. It's fascinating, because it does solve a lot of problems that haven't been solved yet.
Elissa Davis:
Yeah.
Ramia Lloyd:
It's adorable.
Elissa Davis:
Well, and we have all this waste material from construction. Why not use it to-
Ramia Lloyd:
Construct new things.
Elissa Davis:
I mean, it's another method of recycling that doesn't put the ... No, I was going to say, it puts the burden on the companies rather than the individual, but-
Bejamin Moses:
Steve?
Stephen LaMarca:
Yes, sir.
Bejamin Moses:
Let's get some bad news.
Stephen LaMarca:
The bad news. Okay, so for us cutlery and EDC pocket knife enthusiasts, a household company name is Crucible. You've probably seen it before, with CPM, Crucible Particle Metallurgy, before a grade of a stainless steel alloy or a specific blend recipe for a stainless steel alloy that goes on your grossly overpriced pocket knife. Crucible, a while back, I forget if it was either earlier this year or last year, maybe it was before. Crucible is that big company that makes all of these amazing, incredibly nerdy alloys specifically for high-end cutlery. Was purchased by Carpenter, which was a great move on Carpenter's part. I think it was, I thought it was really going to help support Crucible doing what they do best, which is make these really nerdy pocket knife steel alloys. Carpenter, as we know, they are an AMT member, I believe, but they advertise a lot in the additive industry and manufacturing industry in general. They're a very awesome alloy steel powerhouse, and they bought Crucible.
Crucible, sadly, is going out of business and they are starting to liquidate their assets and IP. This is really sad to the pocket knife and cutlery community, because one of the concerns is, well, what's going to happen to these steels? Does that mean the stock of these steel powders is just going to run out, and once it's out, it's gone forever, or is Carpenter going to pick up the slack? Which is probably going to happen, but there is a chance that we could run out of these steel powder alloys, and as a reaction to this, because it's happening faster than the community thought it would, I did end up buying myself a Christmas present over the break, and I got myself a knife made from CPM Magna Cut, which is Crucible's magnum opus blade steel, which has insane edge retention, has insane toughness, meaning it is very unlikely to yield to bending or fracturing when you apply force to it. It's incredibly corrosion resistant and temperature resistant, and holds an edge really well. It's a perfect alloy. Usually, the more perfect the alloy gets, the more of the pain in the butt they are to sharpen, but if you make it good enough, you'll never need to sharpen it.
That's not entirely true, but it'll have a working edge for a very long time. I got one and it's kind of sad, it's the future of Crucible and all of these alloys that Ben and I probably have in our pocket right now, is in limbo.
Bejamin Moses:
This kicks off an interesting scenario, where we've been talking about designer materials, so creating custom materials for different applications. The whole universe of, instead of just buying 300 series stainless steel or 17-4 or inconel, coming up with custom applications and then using those in very unique cases. What happens when the company goes out of business, right?
Stephen LaMarca:
Right.
Bejamin Moses:
The end of life scenario. I mean, we've seen this scenario occur in defense a lot, where, there's an old story about tail hooks for one of the fighters or something that was being designed in the Air Force, where I think they were retrofitting an aircraft to be launched out of a carrier. They said, "We only need 600 tail hooks for this thing," so they had a custom material, said it worked great. 10,000 aircraft later, "I don't know what to do with this thing. Where do we source replacements? How do we keep building this thing?" I think additive and custom materials is going to run into warranty and sustainment issues. How do we handle this at end of life? I mean, luckily for blades, I mean, no one's launching aircraft for this thing, but ...
Stephen LaMarca:
Well, I mean, speaking from the knife community perspective, when you get really nerdy and dumb and spend way too much money on a single pocket knife just to open cardboard boxes on regular basis, instead of actually going out and camping and batoning wood sticks to make fires, when you spend that much money with a lot of these great knife companies that use these insanely impressive steels, if something happens to it, like you actually end up breaking a blade, usually you can message, call the company, and they will replace the blade. You have to send your knife to them and they will disassemble it, put a new blade in it, sharpen it up, true up the action, and send it back to you for a small fee, much less than the initial price of the blade, for certain. What happens when you've got a CPM Magna Cut blade and Crucible is out of business, and you go back to the company that made your knife, whether it's Benchmade or Spyderco or somebody, "I need a new blade for this." It's like, well, it's not going to be the same grade of steel. It's scary.
Bejamin Moses:
It's interesting.
Elissa Davis:
You know this economy, man. First, Party City.
Stephen LaMarca:
I said it's scary. It's not really that scary.
Elissa Davis:
Now, Crucible.
Ramia Lloyd:
Party City, man, that hurts.
Elissa Davis:
Party City has been a staple in my life for so long.
Bejamin Moses:
What happened to Party City?
Ramia Lloyd:
They're shutting down.
Elissa Davis:
They're going out of business.
Stephen LaMarca:
They're taken over by Spirit Halloween?
Elissa Davis:
I don't know, but yeah, they're closing all their stores.
Bejamin Moses:
Wow, RIP.
Elissa Davis:
Yeah, I know.
Stephen LaMarca:
Never mind.
Bejamin Moses:
I have some good news. The last article I want to talk about was Nvidia. We had CES a couple of weeks ago, tons of announcements, but the big focus that I've picked up on, it's obvious, where Nvidia and a lot of other companies are shifting towards supporting artificial intelligence, specifically on the hardware side. That was the biggest interesting takeaway for me, was, there are announcements on creating generative physical AI to power autonomous factories and warehouses. We've seen an explosion of hardware dedicated just for the mathematics behind artificial intelligence and machine learning. It's an interesting thing where, on the gaming side, consumer level, they've had predictive frame generation. When you're running a game, if you push the graphics all the way up, your frame rate's going to take a hit, right? You're running 30 frames per second, that's fine, that's running great, but people want 100, 200, 300.
Stephen LaMarca:
Oh, yeah.
Bejamin Moses:
What they've done is using artificial intelligence to predict the next frame ahead, so you can basically duplicate that. Now what they're doing is predicting three frames ahead, so they have individual hardware and chips and silicon to do just those predictions. If you cascade that to the industrial side, they're working on equipment for servers and deployable computers, so someone at home can start developing their own data set and their own artificial intelligence tool. It's a very interesting scenario where everyone's focused on the software side of AI, but really, the infrastructure is really what's going to help get us there, on two layers, right? One is power consumption. It's very intense, still. Microsoft is still going to buy a nuclear power reactor to power their server farm.
Stephen LaMarca:
Good.
Bejamin Moses:
Where I think we're headed is where, the parallel we've seen in most silicon applications, the hardware side of it is, consume a lot of power up front, but then see efficiency gains over the life of the application.
Stephen LaMarca:
Honestly, two things I want to say. Number one, this is great news for gamers.
Bejamin Moses:
Sure.
Stephen LaMarca:
Because this means that if other people want ... Initially, it seems a little scary, because if other people other than gamers want these high-end silicon to do all of this crazy stuff that is now applicable to these various other industries, what's going to happen to the stuff that I want? Is it going to take a hit? It's actually developing it even better. We saw this starting at the beginning of the pandemic with the quantum leap forward gamers went, when Nvidia went from the GTX to the RTX cards. That was quantum, but you paid for it. Now it seems like in the last four years, with RTX 20 series ... Nvidia just announced the 50 series. That's three new generations of cards in four years, in five years, which in hindsight doesn't actually seem like that much, but it is. It's a lot of development in the ... If you look at the history of how GPUs have been, it's usually one every two years, so they've sped up a lot, but the big announcement that I saw also from Nvidia was, they just announced yesterday the 50 series, and the 5070, the entry level card, is going to be $550, which is pretty expensive for an entry level card, but it already outperforms the release last year, RTX 4090,
Bejamin Moses:
Which was $1,500.
Stephen LaMarca:
An entry level card a year later is now surpassing the most premium, even regarded as industrial grade GPU from the year before. It's insane.
Bejamin Moses:
You did see a reduction in physical dimensions of it, so the thermal capability of it was actually reduced. It's a smaller card, less fans on it, so it's a very interesting package. The thing that I want to hear about is, what's the impact to manufacturing technology? Most of the controllers are basically a PC, right?
Stephen LaMarca:
Yeah.
Bejamin Moses:
That require a CPU, requires GPU. What are we doing differently on board manufacturing technology that's going to require this hardware? Because, we've talked about edge computing versus cloud. Cloud is definitely getting more robust. It's easier to scale up, but at some point you still have to do calculations on machine. I'm definitely kind of interested to see how our members are going to take advantage of this to create new technologies or be more efficient at it. With that, Ramia can tell us where people could find us.
Ramia Lloyd:
Amtonline.org/resources, like, share, subscribe.
Bejamin Moses:
Bye everyone.
Stephen LaMarca:
Bing bong.
