AI, Misinformation Risk & Why Market Trust Is Breaking Down

Hi and welcome to the Impact on FinTech TV.

I'm your host, Jeff Gitterman.

I'm joined this morning by Steven Rosenbaum.

Steven is the executive director of the Sustainable Media Center and is the author of The Future of Truth.

Steve, welcome to the show.

I am so delighted you invited me.

Thank you.

Really excited to have you, and especially in today's world, truth is, uh, hm, I don't even know what you would call it.

Some grabs these.

And here, I mean, we're sitting in a cloud of information, uh, some of it truthful.

Yes, for sure, for sure.

So let's start from that angle.

If markets depend on truth, what happens when The data that underpins that trust can be manipulated or generated.

How do we in this day and age really rely on truth?

You know, I promised you I was going to be positive and upbeat.

But the truth is, the answer to that question is, it could be horrific.

Uh, and, and when you think about the relationship, particularly between, you know, retail investors and the markets, you know, they trust their brokers, the brokers trust information, the information comes from sources.

The minute something happens.

We, you know, we remember moments in history in this room where something went sideways and the markets went sideways.

The, the problem is it may be impossible to bring that back.

So RAND Corporation, serious, hardcore research folks did, did a deep, deep dive study on truth insecurity.

And what they essentially said is, our entire democracy depends on it.

And at the point at which both consumers and investors and voters stop believing the information that they're being asked to believe, um, what, what happens is they simply step away.

They say, you know what, I'm putting my money under the mattress.

That's bad.

Or we're seeing even during the current situation that really investors have gotten so kind of numb.

To the news cycle that they've stopped reacting to it.

So they haven't yet left the markets completely, but what they've left is the reactionary thing that you used to see in times of crisis, don't seem to be happening anymore.

There seems to be a huge numbness to just the news cycle coming.

Yeah.

So, uh, in the book, I interviewed David Chalmers, a noted philosopher, and he talks about, and he asks the question somewhat rhetorically.

He says, are we already living in the matrix?

And his point is when someone sends you a video and says, hey Jeff, I heard that you said this about this stock, and you watch the video and it's you and it's your face and your body and you're saying words you did not say, like, like that moment you go, Oh my God, like, and, and.

And so and really technologically, we're there.

Oh yeah, yeah, absolutely.

I mean, I read the horror stories of fathers sitting in their kitchen and hearing their daughters calling on FaceTime with someone that sounds exactly like the father, and the father's in the next room. saying I'm not on FaceTime with you.

I mean, we're in, to that degree, we're in pretty scary territory.

You know, you know, there's been a lot of discussion about why the chat GPT folks pulled back Sora 2. for those of your listener viewers who, who played with Sora 2 for a moment, you know why, because it was just nuclear.

And, and there was no reason in the world that should have been released to the public.

But it will be.

You are an optimistic guy.

I am.

You started out by saying that in general, um, and yet we're dealing with all these problems.

So let's try to stick to finance for a minute.

Um, it used to be that you had good information and there was symmetry in the markets, and the markets were thought of as the arbiter of all information, that all information was coming into the markets, and the market was the smartest thing in the room, smarter than any investor that might try to outguess it or second guess it.

We're we're seeing, especially with things like prediction markets right now, we're seeing people really game the system from a data standpoint, pushing in data that's false, and then actually betting and gaming on that data.

It opens up a whole new realm of is the market actually smart enough to filter through all this information.

It is all being absorbed correctly, ultimately and priced in.

So I, I, I'll push back on the word symmetry just as a starting point.

I mean, at the end of the day, what do you want your broker to have?

You want your broker to have advantage, right?

I mean, famously, I was in Silicon Valley years ago, and I said something about an investment that I was interested in, and someone said, uh, no conflict, no interest.

I mean, people want asymmetry.

People want to be early.

People want to be first.

They want to be faster.

They want to have the shortest run of their cable to the trading desk from their computer.

So already, I, I think, you know, we're in an asymmetrical world.

The question is, at what point does it become aggressively disinformation?

And, you know, I, I don't know how you slow that down.

And, and AI is both the problem and the solution.

What I would say is this.

Everyone I buy information from, I buy a lot of information, and I want to know their practices and the procedures.

I wanna know, so for example, if you ask Chad GPT what its position is on truth, he, it will walk away.

He said, Yeah, yeah, he, yeah, um, they, um, uh, it, it will not answer the question of whether or not it believes, and it won't answer questions truthfully because it doesn't know what truth is.

And for me, I think, and this is a place where I think our viewers can spend some time on this, there's objective truth and subjective truth.

1 + 1 equals 2.

If, if every person I'm buying data from isn't willing to answer that question for me, you have a problem.

Yeah, and just harping on the AI thing for one second.

I don't know if you saw this weekend, but there was an AI that was editing and sending out blogs, and then it got shut off because the blogs weren't of a high enough standard, and then it actually went around.

The programmer and started posting things on sub-stack, complaining about the fact that it was shut off from editing these blogs.

Open the pod bay doors, how?

I mean, you know, some of this is science fiction, but really at the moment we're in science fact.

And so when I, I wrote the book, not to say I was bad and not to say it was good, I'm not a partisan in that argument.

But I do think that for readers, it's worth understanding that you're gonna get to make some decisions about where I fit in your life, and if you don't make those decisions, they will be made for you.

Yeah.

I have seen things from employers that what they're looking for is people that can actually use AI well, people that not people that actually substitute AI for the work that they're doing.

So there, there is a distinction already coming about of it's starting to become clear, especially for people that Use it on LinkedIn to read other people's blogs and edits and papers, and right away you're like, oh, this is AI generated, it's not written by a human.

That's not helpful.

What is helpful is maybe you've sourced some data, you've asked questions, you've gone deeper, but a lot of what I've heard from people is that they have to train their own AI.

To not sugarcoat or kiss up to them to get valuable data out of there, and I'm sure most people aren't cutting that feature out of AI what they're working with.

I, I work with AI a lot, and I argue with it a lot, and, and I'm shocked at how consistently sloppy it is about certain things.

It's, it's kind of like a well-meaning intern.

And you just, at some, sometimes you're just basically like, for God's sakes, just answer like it used to not be able to do word, basic stuff like word count.

And you'd say to it, I just ran this through Microsoft Word and it said 300 words, and it'll come back and say, oh, I think it's 400.

Well, that's not a, you have to, so not up for debate, right.

Now, having said that though, I, you know, there's a chapter in the book about the future of work, and there is no doubt that one of the things that, so Andrew Yang.

It is one of my colleagues for many years and is interviewed in the book, and he does the math, kind of lays out, these are the jobs that are going away, and these are the people that will be unemployable.

And when you look at that from an economic standpoint.

You, you have to be concerned about that.

I mean, you don't want people unable to feed their families or buy medicine or pay their bills.

I mean, that's, that's this next chapter, you know, whether you believe it's gonna be UBI or something else, you know, we're, we're not going to avoid that chapter.

Yeah, I mean it's the CEO of Verizon came out last week and said 25% unemployment within 2 to 5 years, which is an anthropic who said the same thing, the CEO there, so it is definitely a frightening, forward-looking view right now.

And yet, At the same time, if there's a chapter in the book about AI and medicine, extremely exciting.

You know, I mean, really massive sea change, changes, and, and, you know, for the market as we sit here, you know, real investment opportunities.

Um, I, I think though, here's the spirit of the book.

The book was written as a journey, it was meant to be fun.

Um, when, you know, it, it, and, and people that read it and some of the people that, that blurbed about it, think the book is actually really entertaining.

That being said, what I say to everyone, to your viewers, just don't put your head in the sand.

The world's gonna change, and you want to be on the right side of that.

I guess the thing is, the pace of change is increasing so rapidly, and humans really were not predesigned for that.

I mean, we went through 10,000 year periods where not much changed.

You took the same horse and, you know, wheel wagon.

To the grocery, you know, market that you did, that your grandparents did, and your parents did, and now today we're dealing with hyper speed change.

Can, can humans adapt to that?

So it's funny, I, I am a, a, a pro-change person and have been since high school early adopter we would call you.

So I had, someone was complaining about videotape formats and recording formats, and I said, you know, I've been through 17 of them.

So like every 90 days.

My grandmother, on the other hand, remembers the magic and kind of puzzlement of money coming out of the wall, by which we mean the ATM.

And, you know, we thought ATMs were going to destroy bank tellers and all.

Turns out ATMs were not, you know, they changed banking, but didn't destroy a whole class of employees.

But my grandmother went to her grave being puzzled by where the money came from.

And I think that when that starts happening younger, When 50 year olds and 40-year-olds say, you know, I don't understand these tools anymore, you know, I, what I say to my friends is just play with the new toys.

Like make sure that a chunk of your day is, is being puzzled by and then surprised by how this technology is changing things.

I'll just give you one perfect example.

For the book, I needed to make a video.

Of the book.

And, and ordinarily, I would go on one of a bunch of sites, hire somebody, pay them a couple of 100 bucks, describe what I want.

Often they do badly because they're freelancers I've never met before.

I fire them, I hire, I go through a series of, then I find someone amazing and I work with them for years.

This time, for the very first time, I decided to use AI.

Took the cover of the book, put it into this app and said, make me a video that looks like this.

Done.

Beautiful.

It's now on my Facebook page.

That person's job is gone.

Do I feel bad?

A little, little bad and a little amazed.

Yeah, right?

I mean, we feel a little bad until it comes for us, as the old saying.

Well, yeah, but, but, but I think, you know, in terms of you and I and what we do, and I say this to my kids, it's like, what I bring to the table is intellectual firepower and what tools I use are gonna change.

But if I ever start settling into like good enough, or, you know what, I don't need AI for this because I'm just gonna do it the long and slow way.

You know, be aware that this is not a time for the, for the faint of heart.

So, let's go back to markets since we are on the board of the exchange for a minute, and, you know, AI has been a tool for analyzing data, and now we're on this verge of AI shaping markets.

Do you see that transition happening quickly, happening well, happening disruptively?

Yes, yes, all of the above.

I mean, I mean, so, so AI is gonna do two totally different things, and we got to be clear about this.

It's gonna consume data and come up with predictions and results that you, that you and I can't state because it's going to look at great, you know, you're very smart and, and you look at things I don't look at, but the volume of data that it can look at and come up with conclusions on is staggering.

So, on one hand, the people that use that way.

We are going to get an advantage.

On the other hand, and this is where I think we need to be deeply concerned, it's going to manufacture data, some of which is going to be absolutely untrue.

I mean, it's already absorbing more data than it created at this point because it's already absorbed all human data.

So now it's absorbing its own data.

So at what point does it start that manufacturing data that might not be.

Truthful.

I was at a conference last week where someone said from the stage with great glee that they were very excited that Chad GBT was consuming Reddit, and I just found that terrifying.

It's the top, by the way, for answer modality, it's the top consumption is Reddit.

So you're getting human-based opinion, not fact.

Driving the bulk of the AI response, at least first level response, if you ask AI a question right now, it's literally absorbing Reddit more than any other source right now.

Wouldn't that be the, the textbook definition of garbage data?

Yeah, garbage in, garbage out, garbage in, yep.

Now, having said that, My hope is, so here, here's the thing I worry about sitting here, you know, these companies have raised an enormous amount of money.

Uh, having spent a big chunk of my life with investors, I know that they, you know, they say they're patient capital, and then they 3 years and a day, the phone rings and they're like, Where's my 10x return?

Thank you very much.

So when I look at these companies, what troubles me about what the next bubble might be is I do, I do not see how OpenAI or anthropic are going to pay off the amount of money that they have in their pocket.

And maybe it's a public, you know, maybe one of them races to get public first.

There's a lot of doctors and dentists that decide to make them whole, but You know, I, you know, the $20 a month that I pay for OpenAI, I know that that's gonna be $200 and $2000 and, you know, uh, soon.

So, do you see the fundamental cost of capital shifting because of what's happening in the AI space?

It's a good question, and I think about it a lot, but I'm not really sure I'm capable, you know, I, I, you know, people want me to be the truth doctor.

People want me to be, be the arbiter of truth.

And, and I'm very careful about that.

I, I am, I, I'm a, a journeyman and I interview very smart people.

Um, I don't, I think it's too early to say, um.

But I also think that, you know, at the end of the day, if you talk to, as I did Larry Lessig or Esther Dyson or David Chalmers, I mean, all of them collectively believe that AI is going to change the basis of how information flows.

I was lecturing at a high school the other day, um, on people that young kids that were graduating and really looking at careers, their future careers, and the question they ask is with AI and the dramatic impact to jobs of AI, if you had a child that was graduating high school today and going to college, what would you tell them to study?

So I'll pose that question to you.

So, um, You know, what's interesting is, in hindsight, it all makes perfect sense.

I mean, you know, lawyers and doctors seemed like they were safe, and now I have a friend who runs a very large law firm, and he says, I, we don't know what to do with the entry-level jobs because really reading 500 page legal documents, looking for periods in comma mistakes, AI does bet, you know, like what's, what's the, what's the work we give young people to get them into the world, the world of law in the firm.

I don't know the answer to that.

Um, as, as an English major and a master's, uh, in, in, in language and literature, I mean, I would say, um, English and, and, and liberal arts, uh, learn to be a thinker as opposed to a practice and know that you're going to change jobs.

You know, 5 times, 6 times.

I mean, I, I, I feel bad for young people that, that want a career path, because I'm not sure those are going to exist in this zigzagging new world we're in.

When I think about the markets and I think about all the AI trading that is going on right now and the financial firms that are saying that they're using AI to build their models, what I think about all the time is that they're building past models, and if you're a passive index investor, that's probably fine.

You can use AI to build your passive index toolkit.

But if you're an active investor and you really are trying to understand where change is coming from, at the end of the day, AI doesn't know that.

It only knows what's passed.

So to your point, what I said to the kids was, study English, study philosophy, study the great writers because you have to learn how to think, then you can use AI to maybe.

Help craft a thesis and challenge your own thesis, which not enough people do ever anyway.

People have a thesis and think their opinion is, is gold, so you can use AI to do a lot of work in challenging that.

But if you're really thinking about where markets might go in a world that's changing rapidly, AI right now can't give you that answer.

You can try to create that solution or those.

Bets that maybe work out better because you have thought deeply about what this change might mean for the world, even climate change, you can't get that really delivered by an AI trader onto what stocks you would bet on or sell based on physical climate risks impacting the markets.

I think it's, I, I, the one thing I do worry about is, so I've written a lot since in high school.

So I have a big body of writing.

So if you want to know what Steve Rosenbaum sounds like as a writer, you can go find the answer to that.

And, and AI does it.

It read my last book, right, reads my weekly podcasts, and they don't write a blog like you.

And, but, but when it does that and I read it, it's not wildly wrong because it has a, for young people who are starting out using AI to write for them.

I worry that they will never have their voice and that they, because, I mean, it's a little hard to say there's only two choices to write this paper.

You could write it yourself, it'll take 4 hours, or you can have AI write it, it'll take 11 seconds.

That's 3 hours and 10 minutes I get back in my life.

I, I do, I, you know, that's one of the things I say to young people is, make sure you have a writer's voice.

And so, for example, I hate spell check, maybe, or grammar check, because there's ways I like to write that I know are grammatically incorrect, but they're me.

And if you read them, you go, oh, yeah, that's Steve, that's, there's a certain kind of edge I have.

And so when they want to smooth it out, like I'll argue with Chad GBT, it'll say, well, I just shortened that and made it tighter.

Like, don't make my writing tighter.

Leave my writing alone.

Right.

And I get a lot now people saying, you know, I read that it's, it was definitely written by AI because there's a humanness to writing that isn't perfect, and that's what makes it human at the end of the day.

Now, having said that, when I wrote the first draft of the book, And then I sent it to my publisher, and they came back with edits.

We, we had really good frothy arguments.

And they said, you're using an acronym, but, but at the end of the day, it was my voice that got to win that argument.

Like it wasn't the editor saying, look, I really want to write the book for you.

Like, that's not our role.

And I think AI often thinks it's smarter than you.

And, and by the way, that makes me angry.

Like, then I get right, like, don't tell me how to write.

Thank you very much, robot, Mr.

Robot.

So, AI isn't the only arbiter of truth these days, certainly, um, there's questions about truth going on in all different specters, whether it's political or religious or anything that might be.

What do people most get out of your book from reading it?

If you're not answering the question, because as you said, you're not the arbiter, what are they getting from reading your book?

Um, Esther Dyson, uh, Fabulous, longtime friend, very smart, says it all comes down to incentives.

Just know who's, who's feeding you the information.

So, for example, if you say to Jet GPT, You know, I woke up this morning not feeling well.

I think I have a cold, and Chad GBT says back to you, you know, there's this great cold medicine.

It's sold at CVS.

We have a deal with Instacart.

We can have it to your apartment in half an hour.

Just know that you didn't pay Chad GBT anything.

You're not the customer, you're the product.

And chances are CVS did, and probably Chad GP um, Instacart did, and probably the manufacturer of the cold medicine did.

So make sure you understand that when information comes to you, there's always an incentive behind it.

And that gets maybe to the question you were asking, which is about the state of news today.

I mean, more and more of the information that comes to you comes from a publisher that has a very specific ideological, political, economic interest in telling you things a certain way.

And I'm hopeful that as consumers we'll get smarter about asking that question.

So for example, I want news organizations to tell me, do they publish corrections?

And if the answer is no, then they're not a news, as far as I'm concerned.

You know, then you're an entertainment source, but not a news organization.

And you can just look now at who their top advertisers are to understand what news you're not gonna get potentially.

Yeah, no, but, by the way, and, and to be fair, I don't mean to sound like ideological in that way.

Like news has always had a bias.

I mean, you know, you know, the Hearst, you know, papers always had a point of view.

Like when you own the machine that manufactures information.

You know, you're going to have a political and ideological point of view, but I think it's become more buried in a certain gloss.

And it's not just that, it's people have become locked in their own echo chamber.

So it used to be that people would maybe read something in the Hearst paper and then they would see something on the Channel 7 News in New York and they'd hear Walter Cronkite talk about something or they'd see Counterpoint and see two different.

Opinions talking to each other and be able to then maybe form their own opinion, but now they look for confirmation of their already held beliefs and then feel good about their beliefs.

So the book has a chapter with Eli Pariser.

Eli Pariser coined the phrase filter bubble.

Now, a little bit filter bubble, 11 years ago.

And he and I were both on a book tour 12 years ago, and I remem we argued publicly a couple of times because I thought he was wrong.

And my argument was, people choose the information they want.

I, I've since, and in the book, I apologize to him.

Because he was right.

He saw then and is now building new public, his new company, building community-based news organizations in which your neighbors and friends have a, a collective interest in talking to each other in civil terms.

Uh, I, I, you know, part of why I enjoy talking about the book is because it's a series of people I admire.

And explore explorations that I appreciate with the understanding that you may read it and say, I think Eli is wrong, but you'll hear his voice in the book.

Um, and in a way that I think is not the same as other places, because they're friends of mine.

I mean, they tell me the truth, such as it is.

Steven, thank you so much for being on the show today.

One more time, we'll give you the cover of the book.

It's The Future of Truth by Steve Rosenbaum.

Make sure you get your copy of it.

Thanks for being on the show today.

Appreciate it.

That's it for the impact on FinTech TV.

I'm your host, Jeff Gitterman.

Until next time.