Techstrong Gang – September 10, 2024

Hello everybody. I'm Mike Ard. I'm here with Bonnie Schneider today.

He's co-host for today's edition of the Text Drawing Gang. We're gonna be talking about AI and PCs. Then we're gonna have a little chat about hugging face and what they're doing about AI and sustainability.

And finally, we're gonna talk about well, unions and ai. And the new boss is the same as the old boss you're watching. Text Arm gang.

We'll be back in a minute. All right, everybody, we're back and we have a different alignment for the gang today. So we'll start off with, um, Steven Foskett, who you all know very well.

We've been on the gang many times. Steven, good to see you. Nice to be here.

Thanks for inviting me back. All right. And of course, we have John Schwartz joining us from Northern California.

John, how are things in the valley these days? Things are busy this week. We've got the AI Conference.

I know that we've got Dreamforce next week, plus there's another Justice Department lawsuit against Google this time over. Its ad tech business. So that started this week.

All right, John's busy. And finally, we have a new member of the Gang. Anna Whole Award joins us from Austin, where she is a growth scientist helping companies grow through data.

And welcome to show. Thank you so much. It's wonderful to be here.

All right, let's get this puppy started. ai, which, uh, is contributed from your squad over at Tech Field Day, and it's about, uh, in a report, I guess you guys did during an event about AI PCs and what's going on with this whole thing. And, um, what's your assessment?

What's your take at the highest level? And then we'll dive in. Yeah, I think that this is a pretty positive move, uh, overall.

One of the things that I hear constantly when people are criticizing AI is the whole burn down the rainforest, or it uses as much power as Denmark, or pick a country, you know, that kind of thing. And, you know, honestly, that's true. Uh, but it doesn't have to be true.

And, um, and frankly, it's not gonna be true going forward. So we've already seen Apple have a special AI event where they talked about running on device on these guys. And, um, the same thing happened, uh, on the Windows side with the, uh, a I PC and copilot plus PC announcements with Microsoft, uh, companies like, uh, Qualcomm, uh, a MD and now, actually recently, Intel are delivering low powered devices with, uh, special co-processor that will allow us to do a lot of the AI processing on device instead of in the, uh, in the cloud or in the data center.

That obviously saves a lot of, uh, power. It also saves a lot of money, makes things more responsive, and, uh, the PC makers hope will drive sales because of course, that's really what they're interested in. So I, as you mentioned, I gave a talk recently on what is an A IPC and my premise, my hypothesis is quite simply that an A IPC is hardware plus software, and hopefully, um, maybe, uh, we'll be able to have the kind of software that we need to make these things a reality.

That's certainly what Apple and Microsoft have promised for their various mobile systems, and it's what companies like Google and Samsung and Qualcomm and Intel and everyone are trying to deliver across the, uh, spectrum. So we'll see what happens with this. But, um, you know, I'm optimistic that IPCs will indeed start turning the corner from AI being that sort of big supercomputer, oh my gosh, what do they think of next to burn things down and start getting real and getting functional on an individual level.

All right. And you're new to the show. Are you excited about these devices?

I mean, or is it to you kind of like old hat? Well, I mean, I think the exciting aspect of it is that it's supposed to be personalized. So in theory, this software's gonna learn what I want, what I need, and I mean, that sounds pretty appealing to me.

Anyone who's set up a computer recently knows there's nothing custom about it. So if we can bypass that and get a better user experience, sure, I'll, I'll fork out the three grand for it. Um, but what I find fascinating is that we're acting like this is so new when really it's, it's kind of been around for a minute.

And I mean, I think if you listen to Nvidia, they think that they invented it back in 20, uh, 2018 with the RTX tensor cores. But, you know, I, I'm all for it and I'm definitely gonna be buying one. Steve and I have a question, since you and I have both been around the block more often than we care to admit, but, um, I kind of look at this whole thing, and I laugh a little bit because for so long now we've talked about, well, we need to move all the data into the cloud, and we were gonna offload processing from the end points into the cloud.

And now I look at all this and we're talking about offloading from the cloud to the end point, and it's kind of like the exact reverse of what we've been saying for the last 10 years. And there was a time in my life when I remember that people would say, you know, bring the compute to the data, and we spent a decade bringing data to compute, and now we seem to be coming back full circle. So, um, does this strike you as a little ironic?

Well, there's a couple things to unpack there. Certainly, uh, number one, as you said, there's this idea of offloading data into the cloud for that perspective, I look at AI really as, um, or especially large language models as an interface, a user interface more than anything, essentially, it's a way for us to interact with computers in a new way, and user interfaces run locally, uh, universally. That's basically what modern devices are.

In fact, you know, I think if you look at your laptop or your phone, what you'll find is that that UI code is running locally and the data and processing may happen locally or may happen remotely depending on what happens. But essentially, UIs run locally for reasons of latency and distribution and reliability and performance. And, and so it makes sense to be running an LLM or something like that, uh, locally on your phone.

The second aspect, of course, is the data aspect, which you mentioned as well. And there I think it matters, you know, data locality matters. So if you're trying to coordinate data from a variety of applications that are actually literally in your hand in terms of location of data, then it makes sense maybe to run that literally in your hand.

Whereas if you're trying to do something that requires out of device, um, access, then it makes sense to run that in the cloud. And if you look, uh, and I don't wanna sound too much like an Apple fanboy here, but I really think that Apple's approach is smart because number one, they've devolved AI into numerous different locations and applications on the system. And it's not that AI is an application, though, of course, they do have an image playground thing that nobody cares about it.

It's that AI is integrated in various ways into the operating system and into your workflow. And so that makes a lot of sense. The other thing that Apple is doing is their consciously differentiating between essentially, uh, on device processing and in the cloud processing with regard to data.

So if you are asking the s word, uh, to schedule a a, an appointment or summarize an email or something that happens on device, because that's where that stuff is, if you're asking it, you know, summarize, uh, an article on the web or tell me more about, uh, who's eating the fruits in my garden, then it's gonna go off to OpenAI and chat GPT, which exists off device. And so there's a, a very conscious split there in terms of location of data and locality of processing. So yes, I see what you're saying, but in this case, I think that, that there's some nuance to it that's important to consider.

You know, Mike, um, it's, it's interesting that, and I can ask Babe Ann or Steven about this, the impact of what Apple's gonna do, what they've announced on Monday, I mean, Dell, Dell Technologies world, they talked about, or Dell projected this, uh, super cycle of a IPCs and, uh, kind of in a sense, saving the business or buying business sector where sales are gonna be on a steady, have been on a steady decline since the pandemic. This is going to basically reverse things. So I think as far as I know, there's Dell, Samsung, HPE, Lenovo, Acer at least, have unveiled their own brands of copilot plus computers with Apple.

One of the things that's interesting to me is not only are they getting into the market where in a sense they establish a market and accelerate the sales in that market, but they've, there are 300, I believe there are 300 million iPhones worldwide that have not been upgraded in more than four years. And I think that the idea of the concept is that is actually going to spur and open the floodgates. So I kinda wanted to throw that out to either Steven or Mike or Bonnie or or Ann about the, what the impact on Apple will be on ai.

I think that they were very slow to start talking about it, and I think there's a reason they have a tendency, if they're not the, the first mover, they have a tendency to, uh, take their time. We saw that with the Apple Vision Pro, but we all know that they're not betting the form on that. Uh, the sales were, were lackluster.

So I think, you know, planned obsolescence is something that I think they've been accused of for years with the iPhone. And so, you know, they make a good product so it lasts. So I, I'm not surprised to see that those numbers are stagnating.

So of course, they have to reinvent themselves just to get that extra money out of us. I don't know, I'm a little dubious of the timeline. I don't think that this is a, a, a hockey sticks by thing.

I think it's more of a longer term upgrade cycle. And the reason of this is, I think it's gonna take a little time for developers and the folks that they're working with on the data science side that wrap their heads around one of my processing where and when, and build that into an application that creates something that's compelling. But Steven, I know maybe you think that it's more of a hockey stick, or is it a long term kind of bell curvy thing?

Well, I think it's going to, I think it's gonna surprise us by how rapidly the, uh, client side adoption of AI applications happens. I think, um, on the Apple ecosystem certainly, but also on the Windows ecosystem, uh, I, I think that we're gonna be surprised at how compelling these things are. Certainly, it's gonna be a little longer than Apple or Microsoft would want.

I mean, on the Apple case, uh, you know, they're not even releasing all of Apple Intelligence right away. They're gonna be releasing new phones that can support it. But, uh, and new laptops and so on in couple months theoretically, but frankly, those things aren't gonna be, the software's not gonna be ready for months or maybe even another year.

And so that is going to delay this. But I think that once people get exposed to this, they, it is going to kick off a rapid upgrade cycle. And in terms of, uh, planned obsolescence, let me just point out, I mean, apple made a big mistake in that they didn't equip their phones with enough RAM until recently.

And that shows that they really weren't ready for this change. It is, uh, pretty obvious that the gating factor toward Apple intelligence is not the performance of the silicon, which is actually pretty good, even in the M1, you know, an iPhone 14 kind of generation. It's the amount of RAM available to process this data.

And frankly, unfortunately, that's gonna mean that a lot of people are gonna be griping when they find that their iPhone that they bought last year or maybe even this year, doesn't support these new features. 'cause once they see them, I think they're gonna be excited by them. And it's the same honestly, on the Windows side, where once people start seeing what, uh, these new surface systems can do, uh, I think they're gonna want them, but at the same time, uh, they're gonna need a new machine to do it, because frankly, you just don't have the compute power in yesterday's pc.

And that's kind of surprising. So whenever Apple announces something, there's this palpable excitement regardless of the merits of the news. But I think, and also the, the, the thing that the, I guess the argument that's been made, or the topic that's just coming up or the issue is that Apple waited too long to get into ai, but, uh, you necessarily think that's true.

Now, I, I, I don't, I don't really think it applies as much to AI as it did to previous markets that Apple jumped into and established. I think Apple was smart to hold off in jumping in with both feet, but I would really be interested in what Anne thinks. I absolutely don't think they were too late for anything.

I think that it was a measured approach, which they often take. I do think that it's disappointing the right to repair movement didn't do more than it did because we can't upgrade these phones. In theory, we should be able to add memory to our phones.

You know, I remember installing memory in, in college in my, in my, uh, you know, Tangerine iBook, but you, you can't do that anymore. And, and so of course, that forces us to buy new hardware, and that I don't think was accidental by any means. I agree.

I, I did a, a report on the right to repair and apple's, uh, initial reluctance and now starting to, uh, be a little bit more forward thinking with it. But, um, that's definitely an issue in terms of the hardware and then extending the life of the hardware. And what do you, what do you do towards the end of it?

All right, I have an important question. It is September and I might be thinking about what's going on that holiday list at the end of the year. So am I putting new gear on it for this year, or am I gonna wait till next year?

Steven, I think you're putting new gear gear on it for this year. I think a lot of us, well, especially us, are gonna upgrade. All right, well, folks, you heard it here.

Santa says, we're getting new hardware this year, and it's gonna be all AI driven. Cross your fingers and hopefully the software won't be too far behind. We'll be back in a minute.

Discover Techron Group, the epicenter of tech innovation. We are your go-to for reaching IT leaders and practitioners worldwide. Our secret impactful content that sparks awareness, engagement, and top quality leads with us.

You'll access editorial websites, streaming videos, virtual events, custom content analyst research, and more. Join our satisfied clients. Let's revolutionize your tech journey.

Contact us today and tell your story to the world in the most powerful way with Techstrong Group. Welcome back to the Techstrong Gang. 2 billion.

And they doubled their value just in the past few years due to investments from other major tech companies. It was a, a fascinating conversation where we talked about their open source community and of course their sustainable approach to ai, specifically with their many models. Let's take a listen.

This is a preview from the interview, Thomas, it's so great to have you here. Hi, Bernie. Yeah, happy to be here.

Well, tell us about your background and your journey with starting this company. I Was first a researcher in, in physics, quantum physics. I worked a bit in the US at the time in the Bay Area.

I, I joined force with Clem and Julia, my co-founders, to create hugging face. And in the beginning, hugging face was a game company. So that's why we have this kind of unserious name that we are, that we have to have, I think today, which contrasts maybe with many companies in the field.

And we pivoted this idea of building a platform for open source, open science, open knowledge and ai. Can you explain the models that you're introducing and why you think this is part of a, a bigger picture for sustainability? Yeah, I think in many case, I mean, it sounds quite obvious, but surprisingly we still have to say it.

You don't need a model that can solve very complex mathematical problems. I don't know the remain conjecture to just help you in your daily automated task. I think it's something that's not pushed a lot by close source providers because they mostly would like to simplify their stack, and also they want you to pay by token.

But if you have a small model, so first it's quite cheaper to run. And the second is that you can even have this model run locally. So instead of, you know, having this model running on, on data centers, uh, a set of models that you released last month is, is called a small lm, which is set of, of much, much smaller model that you can try on the ha iFace hub.

And they work really well for many tasks and they can run in browser. So you don't even have to send your data out of your computer, it can actually just directly run inside your browser. It's small enough, it's smaller than a, than all.

They always stream from Instagram, for instance, like a couple of hundred megabytes. Basically. I think this product can be really useful for many tasks and for all of these tasks that we automate with smaller model, the energy that's required is much smaller.

What Has been the reception for these smaller models from the people that you're talking to? Really great. Yeah.

A lot of excitement. A lot of the reason we didn't use them is that people thought it was not possible. It was not possible to have small model with interesting capabilities.

And that's the discovery of the past 10 months maybe, or at least at Target phase the past eight months, is to apply some of the learning from the, the large model. If you adapt some of this learning, you can actually have this model with very interesting capabilities. You Started off by saying that hugging face is such a unique name, it's so warm and friendly, but it started off because you were in gaming.

Are you happy that you kept the name now that the company's evolved? Yeah, very happy. I think it's also a good reminder for us that whatever we do, we should not take ourself too seriously as well.

I mean, it's also nice to enjoy what we do, and that's one of the reason I feel very lucky to be here, even after some time in the field that I still very excited about the project we work on. I still find them really interesting. Maybe one lesson for people who are in ai, it's still quite a fun field.

What we do is really something amazing when you think about it as a little kid, right? You can have this machine that talk to you like they are almost like little characters and the, and we keep this like freshness of being amazed, uh, by AI and not take this too, too seriously. At least that's what we try to do at Target face.

Yeah, that's Great. Well, Thomas, it's been a pleasure talking to you, co-founder, chief science officer at Aase. Thanks so much for joining me.

Thanks Bernie. So that was just a little taste of the whole interview. You can catch the full interview on Text Strong TV where it is posted online.

And it was really interesting to speak with Thomas because I was thinking when I first started talking to him, well, this must be such a huge company with the evaluation that that they have now, but he told me that they make an effort to keep hugging face right around 200 employees, though their Slack channel because of all their partnership has thousands that are connected to it. So the community continues. He said also the mini models have been extremely successful.

So they are going to see more of that in the future with smaller models. And as Steven mentioned in our earlier segment, um, doing that are more sustainable because they are doing a lot of processing on location and they're also just leaning into less state overall and they found them to be successful. So we'll be seeing more partnerships evolve with hiking face.

Alright, and what's your take on all these mini models? And while we're at it, models seem to come in t-shirt sizes now as large, medium and small. And, um, are we gonna get smarter about right sizing these things and maybe not have as big a energy crisis that Steven alluded to where I think the forecast is that these things will consume as much as the entire country of Spain, I think was the last saw.

Yeah, And you make a great point. Um, I would say that running models locally is great. I would say that running models in your browser great, but I don't think it's a one size fits all.

I will tell you that I am anti degrowth, which means I think that we should be working on the bigger problem of, of, you know, nuclear energy, solar power. I don't think we should hold back mm-hmm, uh, to hold back energy. I think that's failing in, in climbing up the kjv scale, which I think is something we should aspire to as a society.

I think you could judge a society as Kjv stated by their efficiency of their use of computational power. I think that we shouldn't limit innovation just to be more ecological. We should be more ecological in general and not hold ourselves back.

Yeah, I mean, I think that I agree. I mean, I think there's a place for smaller models and there's a place for larger models as well. And I'm glad that you mentioned the energy sources because just not just speaking to Thomas, but speaking to other, um, leaders in this space, they, they're all leading into renewables as sources and, and trying to be more innovative, even with nuclear energy as well.

So you're gonna see continued investment in solar and wind to power these data centers as they're just getting built at a rapid pace and the demand will continue to grow. Steven, I was just reading over the weekend how, uh, dominion Energy in Virginia's gonna be adding two transmission lines to all create the largest data center corridor in the world. But at the same time, they also noted that, um, it, the waiting list to bring a new data center online is now seven years.

Um, so will we get to this in time for this AI explosion or are we gonna have this, uh, massive, uh, supply issue in demand crisis? Well, I think that in the short term, we may indeed have a demand crisis as more and more electricity is being sucked up by data centers and electric vehicles and electrification generally, and, you know, decarbonization. Uh, but in the long term, I think there's actually really, really great news.

And that's the incredible learning curve of solar panel production and, uh, incredible cost reduction of solar panels. A friend of mine, uh, Casey, er, uh, Terraform Industries, uh, if you want to Google him, has a great blog where he writes a lot about the impact of, uh, the reduction in cost, the rapid, rapid reduction in cost of solar panels. And another friend of mine online recently, um, extrapolated from that, that we could actually get into a perverse situation where solar electricity becomes so cheap that the, um, uh, associated electronics to get it hooked to the grid become impractical.

And we just roll out solar everywhere for just, you know, and just let it let it go to waste when we're not using it because solar electricity is gonna be so, so cheap. Now, I'm not sure if it's gonna get quite that cheap, but it is definitely getting much, much cheaper every single year to the, to the state, you know, that that, that I think that we're going to have plenty of electric energy from solar very, very soon, probably sooner than people know. And, you know, we talked a little bit about the hockey stick here.

I think that's gonna be another hockey stick that we're facing a positive one in, in the, in short order. But I really enjoyed this interview, Bonnie. I I love what you had to see.

I mean, it really jives with what we were talking about in the first segment and what we're hearing about at AI Field Day, which I have to point out is next week as well. That's great. You know, I, I agree.

I think the solar, I, I agree on the rate of growth, and I think that's also coming from more innovative technologies on how to store solar energy and how to keep it, because the initial argument was, well, it's not sunny every day everywhere. How are we gonna rely on solar energy? So as the technology improves to store, um, we'll see more of that.

And you're also seeing, you mentioned dominion, that there's so many companies up in the northeast, Mike and the New York area that are developing new wind and offshore energy that we're seeing, um, with wind, the development every day, um, just, um, offshore is growing, uh, uh, in Texas, for example, and in off the coast of New England. So we're gonna see a lot more of that as all this is happening at the same time. Yeah.

Uh, the report that I read actually noted that, um, Austin and Houston have the biggest data center growth going, but John, I wanted to get your opinion about this. There's also lawsuits being filed surprisingly, um, people who live in these areas say that the data centers are ugly and they're ruining the landscape, and it's just kind of a, generally speaking, uh, uh, the Yes DYS dystopian hell was the phrase dystopian Actually the lawsuit. Yeah, that's a great way to describe it.

So I, you know, I actually, this make me, make me sound as if I'm strange if you don't think already, but there are, uh, I, I like going to data centers. I like touring them. I've toured many of them.

There was the one, there was one that, um, at and t has somewhere in Virginia. They literally blindfolded me to get to it several years ago. They have a moat around it.

They have security guards who are armed. They, it, it's basically a fortress. It's billions of dollars of equipment and it's a secret location, et cetera, et cetera.

They are ugly, but I find them fascinating. There's also one that was being built outside of Kansas City for meta a couple of years ago where I began to see it. What they did was they had it on the outskirts, uh, near the airport, far away.

It's kind of desolate, but yes, it isn't. They are eye source. But I find them absolutely fascinating.

I wish we didn't build as many as we are, but nonetheless, um, you're, I can see a lot of lawsuits, uh, in in ingestion to them, especially in urban areas. So you're saying these things are so ugly that they're just can, you can't take your eyes off Well, from the inside. Yeah.

Maybe the outside there, they're like, not, not so much it look like prisons basically to me at times. But inside, there're actually Someone interesting. I think they should lean, lean into the, um, sustainability quality of, of what they're trying to create and maybe bring that into the aesthetics so they're not, so yes, pri Yes, that would help.

Hey, you know what I was gonna ask Bonnie. It was interesting. The interviews are interesting and I, I, I kept thinking when you were talking to Mr.

Wolf, this company's been around, I think about 11 years. Yeah, I think you mentioned that. And I think it was seven or eight years.

He said, okay, It, it, it, so it's been around nearly a decade. It's making this transition from gaming to a key ai. You know, it was interesting.

It reminded me of another company right out here in California. I'm thinking of Nvidia. It's this kind of whole kind of parallel arc, although NVIDIA's been around longer.

But anyway, I just found it fascinating what they were up to, and it was very exciting. Uh, what they're gonna try to do based on the interview you had with him. Thanks.

I mean, it's, it is, it's kind of cute hugging face. That's what caught my eye when I originally forgot to him, because I, I saw the, the, the logo and I, and, uh, I thought it was interesting that it was chatbot gaming to start, and now he says they're going to, they keeping the name because it, it matches, uh, the, the essence of the company. So I guess it started out one way and then it ended up in a way that all supports it.

All right, Steven, are GPUs gonna get more energy efficient? Can we run things on things other than GPUs that are more efficient? We see all these AI accelerators, our friends of Tuum have a report about this market that's out talking about that.

There'll be, um, other processors in play for ai. So my question to you is, uh, is it possible we might be over provisioning our data centers because between solar panels and everything else that gets more energy efficient, maybe we don't need as many of these big ugly boxes. I think we do need some big ugly boxes in a lot of GPUs, um, for training and building models, basically the software development side of things.

But ultimately the processing as, as discussed on this interview as discussed, uh, again and again, uh, the processing, uh, the client applications of these models are gonna be local and on different devices. So certainly we are seeing a proliferation of silicon that can support AI applications beyond GPUs. Uh, Nvidia and a MD and Intel to some extent have, uh, tried to, uh, I guess capitalize on the demand for GPUs, but ultimately it's going to be special purpose npu neural processing units that are gonna be running a lot of this stuff.

That's what Apple has really focused on in their Apple silicon product. It's what Qualcomm has rolled out in, um, in their Snap Track and X series. And, uh, we're seeing a lot more of that as well.

I'm seeing some really exciting motion happening in the risk five world of adding specific neural processing, um, language, uh, possibilities, right di directly into open source CPUs. We're hearing about that from arm. We're seeing Intel do that as well.

And I do think that there's gonna be a proliferation of processor types and they'll be specifically designed to do this. So we saw an announcement recently where, um, Microsoft and Amazon have developed specific, um, processor silicon just for the kind of models that they're running there. Microsoft has a processor designed to run chat, GPT.

So basically chat GPT specific silicon. Google has their incredible, um, you know, tensor processor unit that they're using in the data center. And apparently that's what Apple used to develop their models.

And I think that that again, is the, the direction we're gonna head. We're gonna see companies using, um, big, uh, AI supercomputers, for lack of a better term in the cloud to develop these models. They're gonna refine them, they're going to make smaller versions of them, they're gonna distill them, and then they're gonna put them out and they're gonna run on, um, special purpose devices.

That's why I think that the boom in AI silicon is actually just starting with GPUs. Soon, we're gonna see a huge boom in client silicon. It hasn't happened yet, but I think it's happening now.

All right, folks, you heard it here. Hey, on the downside, we need more energy and we gotta figure out how ways to build those data centers. But hey, the good news is we're getting smarter about how we consume all that energy.

We'll be back in a minute. All right, folks, we're back. And we're talking about, well, AI and labor.

Apparently, uh, unions are fighting themselves negotiating more often with, uh, organizations that are using AI to figure out how to maybe optimize their tactics, we shall say. But John, this is your story. Walk us through.

Okay, so you mentioned that we've had some labor negotiations over the last year involving, of course, the famous Screen Actors Guild, American Federation of Television and radio artists. Well, what a mouthful. That is Writers Guild of America called Sagara Sagara.

Yeah, Sag, I know, I just had to be official. We had the casino workers, hospitality workers in Las Vegas casino workers. Now, the one interesting example that I came across that actually spurred me to do the story is the International Longshoreman's Association, which is a pretty hardcore group.

They, uh, they have some members in the east and Gulf Coast, so word technology's gonna jeopardize their jobs, their high paying jobs, up, up to about, guess about a hundred thousand dollars. Specifically, the beef is over an automated gate at a small port in Mobile, Alabama, which has prompted the president to say, we will never allow automation to come into our union and try to put us out of work as long as I'm alive. So we're starting to see now union reps expecting more disputes in the coming years as AI improves and is deployed throughout the workforce to improve operational efficiency, cut costs, and replace some humans.

I think the, the biggest concern among the unions is, uh, as a threat to jobs in manufacturing in the form of bots, which we've already been seeing for years with Amazon, for instance, and the distribution of goods, which is interesting because self-serving vehicles may be replacing humans. So we've got all this going on, uh, and in a sense, it's probably not going to get much easier. The negotiations are becoming longer and more drawn out.

There's one thing that I'll mention as the, uh, the casino workers in Las Vegas who were represented by the Culinary Workers Union, reached a new contract late last year that included severance of about $2,000 for each year. The employee worked, if the employee's role was eliminated because of technology or ai. There's also the hospitality workers at Caesar's Entertainment in Las Vegas last year on a provision in their contract with six months notice in the event Advent AI robotics or any other technology was going to be deployed.

So they needed to know, so they could either re-skill look for another job or, or what have you. It's, it's, I think it's very interesting and I think it's gonna become even more heated across all types of unionized jobs. And do we need to think about this in a more holistic way?

Because I feel like these are all symptoms of a larger issue, and we're gonna have to re-skill the workforce, but I feel like we're not driving that conversation as aggressively as we should be. I think that it's a normal human tendency to be fearful, and I think fear holds us back. I always like to look backwards before we look forwards, which is holistic.

Uh, the Victorians were terrified of gaslighting, even though gaslighting meant that they could read better at night versus candlelight. There were costs that were paid for that innovation. People died of, uh, you know, PO air poisoning, basically monoxide poisoning because there was too much gas in the room.

But that advancement led to others. We are not shedding tears right now for the elevator operator or the people that, you know, commanded the horses of the horse drawn buggies we're, we're not shedding tears for them. And there's a reason.

And that reason is that human innovation must move forward. Human innovation is going to include ai. We know this, but not everyone is accepting it.

It's a matter of acceptance and it's a matter of education. We have to educate people as to why this makes them super human. Why this does not hold you back.

It makes you better. We can take repetitive processes, we can automate them. We can actually free up your human time, and it's gonna be maybe be generational before we see that excitement and adoption, but we shouldn't allow ourselves to be held back, uh, because of fear, because that's all it is.

I, I agree. I think if you look at the SAG AFTRA agreement regarding ai, that was, um, foundational in 2023. One of the things that I think obviously that motivated were all these voice actors and, and animators that were saying, we're gonna be replaced by ai.

So the way that they were able to, at least for now because it's only a year later, is, um, say, well, we're not gonna use AI or put your voice into a large language learning model. Without your permission, you're going to be aware, but we're not gonna do it behind your back. And that is what they agreed to.

But what, what I think, uh, I was actually a member of SAG aft up until recently, and what what I think is interesting is, okay, that's fine. I could see that for the big, big actors that we all know, and maybe we'd catch it if it was in a commercial, but if you have others that are doing, um, that sort of work, it seems like it would be hard to enforce. I guess that Would be able to That's, that's a very good point, Bonnie, because, uh, at my previous job I did a whole story on this, this topic.

And unless you were a well-known voice, or very well established as a voice celebrity, like a Bradley Cooper, I talked to a number of people who lost their jobs almost overnight. Their, their, their voices were replicated by the company they were working with, the association they were working with, and they had no recourse at the time. You know, one thing, um, also Anne mentioned, which is interesting, is in California, uh, Newsom just announced a partnership with Nvidia, uh, training state employees, teachers, and students.

So there is an effort, at least in the state to start reskilling or upskilling, which is a major problem, right, among employers. Well, oh, go ahead. Absolutely, absolutely agree.

And I, I think that's important. California, I think we can expect is gonna lead the way there, uh, given how much financially is at stake. But I think that we have a tendency in our society and just our viewpoints now to assume how it is today is how it's going to be.

So I think we need to do a better job as technologists and futurists of painting a future for people. That doesn't mean that they're no longer necessary. It means that, that they're going to have super human capabilities.

And I think if we can shift that thinking, we're gonna see so much, uh, less resistance. Right. But it's also Part, I think employees Yeah, just one, I'll just be really quick.

Sorry Steven. A lot of employees are open to the idea of this. It's just that there is a kind of a disconnect between the, the top level and then the rank and file in terms of organizing these, these types of things.

I, it is gonna happen. Um, it's inevitable. Yeah.

If you, Stephen, to John's point, I don't think I can recall any technology advance that was held up because of some bargaining agreement with unions. I mean, from robotics and manufacturing plants to, um, just this weekend, I think I saw that somebody was using AI to drive window washing without people on the side of buildings, which is probably a good thing. 'cause, you know, every six months ago, we gotta like rescue somebody from one of those things.

The sky high robotics, actually, skyline Robotics, I'm sorry, we just destroyed that. That was the first in New York. That is the first example.

It's on Avenue of the Americas. Mm-Hmm. Right.

It's a bomb squad. We're not, we're not crying about the robots going in there first. There you go.

So Steven, do you think that any of these negotiations are gonna hold up anything? Or do you think this is just kind of, uh, you know, it's all over and done with except for the screaming part? Well, as I was gonna say, if you've studied this stuff, uh, technological process or, uh, progression is obviously inevitable.

And, uh, resistance to technological progression is also inevitable. Uh, you know, we, we call people Luddites because of a specific uprising against specific technological progress. Um, you know, economists will tell us that if you, uh, redu, if you increase automation, that it reduces, uh, supply costs, increases quality, and requires an increase in, uh, labor inputs, which is what the history of the world has shown us.

But unfortunately, that's faint, uh, hope for people whose jobs are lost and who are deflected by those, uh, technological advancements. You know, there was a time when a lot more people were employed in a lot more industries. Uh, you know, I live in Ohio, which was a big auto manufacturing, um, area.

Uh, car plants are really wonderfully productive today here in Ohio and in Michigan and places like this, but they don't employ as nearly as many workers as they used to because of automation. And you'll hear a lot of local folks here saying that they miss, uh, the jobs at the factory, which maybe weren't great jobs to begin with, but they were my job and I miss having my job. I think that's exactly what we're looking at here.

Uh, sure, uh, these negotiations aren't in the long haul going to derail, uh, AI or any kind of technological advancements. But, uh, you know, it's awful hard for people to adjust as these things happen. And as we've seen, for example, in the coal industry, it's hard to retrain people.

It's hard to get them to switch to a different, uh, a different field, even if it is a positive field, especially when we're not exactly sure exactly where they should switch to and what they should train on. But, um, ultimately their children, I think, will benefit from these and their children's children will benefit, uh, quite a lot from the advancements in technology. And so, uh, that's not really gonna reassure you around the coffee table, but frankly, that's how it's gonna be.

And is this gonna become a political issue? Will we see this topic raise its head in presidential debates that might be coming up? Absolutely.

And I would argue it already is, especially some of the legislation on the table of California trying to regulate ai that's not going so well. I mean, I'm, I'm of the belief, maybe let's get section two 30 right, and move on. But, uh, yeah, I think it's gonna continue to be polarizing.

I think it's gonna continue to be political. Um, some people are just gonna be in that camp that fear it, and that's just how it's gonna be until they see a use for it. You know, no one cared about augmented reality until they, they were doing Snapchat features, you know, the doing the funny things on their face.

We haven't really, I think in some ways we've reached mass adoption, but I don't think that people feel that we have. John, how is that Bill going in California Anyway? Well, I was just gonna mention that you've read my mind actually.

Um, so Gavin Newsom is considering it. He's never given any clear indication whether he's going to sign it or reject it. Uh, so we're waiting, he hasn't until September 30th.

Uh, there's a, there is a very large contingent of not just tech companies, but representatives like Nancy Pelosi. So Lofgren, NSU has retired. There are dead set against, against this bill, which would, um, and again, safeguard ai.

It, it's a pretty nebulously written bill, but I think it's probably inevitable. California tends to be the count, the, the state that starts these bills and turns 'em into law and then has a reciprocal effect. I just don't expect any type of national legislation on anything involving tech.

I I've been waiting for that for like a decade, so, um, I'm not gonna wait much longer. I, I've given up on that thought. A lot of the, the, you know, found Godfather godmother of AI came out against the bill, but it reminds me of the publishing bill that California, uh, put out where they said you had to be compensated to show up in Google News or, or these other aggregation sites in Google.

And the results of that was that those news results were hidden. Uh, so I think it could go badly. I Mean, know the only, the only real, there are two names that support the bill, Elon Musk, there you go.

And, uh, Jeffrey Hinton. So beyond those two and a couple of other folks, it's, it's universally been opposed. All right.

This is Where we need Alan to comment, I guess. Oh, I know. Oh, I know the, his regulation.

Well, well, well, let me pretend to be Alan for a minute here. He'll say, you know, because of the dysfunction in Washington, and we have no ability to have any kind of bipartisan agreements, therefore the state of California and Sacramento has now become the default, but leading on all these decisions and everything that's driving politics. And if, if you're okay with that, then, you know, fine, if you're not, you should do something about what's going on in Washington.

How's that? And Again, that's pretty good. Yeah, no, that's, we, I've heard that.

Yeah, we've heard that Mike right? A few times, but I, I, I agree with that. I mean, although I wouldn't refer to these companies, uh, this is a, another episode as national treasures or whatever we call, whatever we call them.

They're just companies. They're, they're just, where they, they just happen to be located in a certain state that has a lot of people. Is that the, that's The, yeah, that's about it.

Yeah. Well, Some of them seem to be, uh, escaping from that state too. So there's that.

There is that too. Well, there's more distribution of people in the places where there's nobody living there anymore because they don't have a job. Right, Steve?

It all comes full circle. All right, folks, I think we're gonna end it there before it gets too much more political. Um, I wanna thank you all for watching the latest episode, and of course, thank you all, our guest, including our newest member.

Ann, thanks for being here. And with that, we're gonna encourage you all to watch the rest of the Textron TV lineup. We got an awesome number of interviews and conversations and content from Futo.

We'll see you next time.