Human Pace vs. AI Acceleration: Growth in the Age of Technology

Listen to the BOLD Business Podcast

       

Search Blogs and Podcasts

Human Pace vs. AI Acceleration: Growth in the Age of Technology

3 Key Takeaways

AI trailblazer Jonathan Schaeffer shares technology’s rapid transformation, the responsibility of leaders, and how to protect your organization’s digital future:

  1. Discover how AI’s ease of use is shifting us from active learners to outcome optimizers, and what that really means for your business or career.
  2. Find out why true responsibility with AI requires us to slow down, challenge the status quo, and set our own human pace for change.
  3. Get the first actionable steps to build digital sovereignty, safeguarding your data and decision-making power in a cloud-enabled world.

What You Will Hear:

01:55 Our current reality is one where rapid technological change is the norm, making it essential to pause and recognize the common ground within that constant change.

  • The major tech developments and the unpredictable nature of the tech landscape.
  • Parallels are drawn between handling technological change and how educational icons approached tough conversations, emphasizing similarity over fear.
  • The value of pausing to reflect, even amidst rapid shifts.

08:25 The challenge of fostering genuine learning in a world where optimization and shortcuts, often enabled by AI, are increasingly prevalent.

  • Academic integrity issues have resulted in a significant portion of students failing due to misuse of AI tools to shortcut assignments.
  • Personal experiences underline a shift from passion in teaching to frustration, as time is spent managing academic dishonesty instead of mentoring.
  • The loss of in-person engagement and motivation highlights how technology is changing teacher-student dynamics.

17:30 There is a growing mismatch between rapidly advancing technology, optimization in life, and the slow pace of societal and educational expectations.

  • Expectations about higher education have become mainstream, making the transition harder in today’s fast-moving world.
  • Changing landscape is prompting new challenges for parents and students figuring out which tools actually help achieve their goals.
  • A hope that increased options will lead to more autonomy for the next generation, despite societal inertia.

20:45 Cameo: Heather Gibson, Owner, Big Hearted Breeders shares how her regular check-in practice can become your biggest competitive advantage in a chaotic world.

23:20 Calls for a technological pause face resistance as market forces and company agendas push relentless AI advancement, despite society’s lack of preparation.

  • Rapid development is compared to an unstoppable tsunami, overwhelming society’s capacity to adapt.
  • The lack of basic public education around AI despite its inevitable and broad integration.
  • The crucial need for increased awareness about AI’s real capabilities and risks, outside of formal education.

29:50 Corporate focus on achieving artificial general intelligence drives aggressive progress, often at the expense of critical concerns like privacy and individual impact.

  • Companies prioritize winning the AI race, downplaying or postponing issues related to hallucination, security, and environmental costs.
  • Short-term fixes are opted for over meaningful responsibility as firms pursue massive financial rewards.
  • Increasing disconnect between corporate incentives and the needs or pace of the average person.

34:45 Human systems and institutions move slowly compared to the exponential growth speed of AI, resulting in society being reactive rather than proactive.

  • The inability to keep up means problems are addressed only after they become visible, not in anticipation.
  • Stakeholders like governments and professional bodies lag, while AI continues to outpace regulation and policy responses.
  • Recognition that forethought and preparation are required now to avoid avoidable crises, yet systemic change is slow.

43:55 It is essential to treat AI as an augmenting tool rather than a replacement for human capability, resisting attempts to anthropomorphize software.

  • Current strategy by tech companies encourages users to see AI as a friend, making users less critical of its limitations or errors.
  • Treating AI as a helper or assistant, not a peer, keeps human users in charge of judgment and decision-making.
  • A warning against attributing emotion or awareness to AI, which remains a fundamentally non-human tool.

52:35 A sustainable path forward with AI requires addressing existing issues first, prioritizing education, and ensuring value for people and companies.

  • Continued innovation without addressing foundational flaws only escalates existing problems.
  • Raising public understanding of AI’s capabilities and limits is necessary to support responsible adoption.
  • Success will be determined by whether AI delivers clear, positive outcomes for users rather than risks and complications.

54:50 It is BOLD to refuse the fear of missing out and actively choose to build an AI platform that is completely disconnected from the cloud.

A RIGOROUS WEEKLY PRACTICE THAT TRANSITIONS YOU FROM THE EXHAUSTION OF BAD GOAL CHASING TO THE CLARITY.

FROM BUSY WORK ➔ STRATEGIC COMPOUNDING

With Red Direction Present Retreat Methodology, reclaim executive cognitive bandwidth to sharpen Business Instinct and “Good Goal” alignment validated by a 40% reduction in decision-cycle time.

Resources

Transcript

Jonathan Schaeffer 00:00
We’re not well suited to keep up with that pace. It’s like technology is, it’s racing, it’s sprinting, and we’re walking.

Announcer 00:15
This is the Bold Business Podcast, created for companies running real revenue. Your host is Strategic Growth Advisor Jess Dewell. Here, we drop the I’m fine mask. Here, we face today’s market demands together. Here, we build business instinct. Because the action we take today creates our company’s future relevance and the room for our teams to grow. Now, grab your tea or coffee and get ready to build your business instinct.

Jess Dewell 00:50
Dr. Jonathan Schaeffer, an AI pioneer, joins me today. He has over 40 years in machine learning and is the founder of Kind by Synsira. As an expert in local data sovereignty, he is building privacy-first desktop AI that puts you, leaders of organizations, back in control of your own information. Listen out for three things in our conversation. The first, AI has allowed optimization, maybe even laziness, to get the desired outcomes instead of learning and thinking for ourselves. Two, being responsible with AI and technology today means it can’t be about the money.
It has to be about taking time to slow down and think because change happens at a human pace. And three, how you can take the first three steps to build a digitally sovereign business responsibly.

Jonathan Schaeffer 01:53
In other words, it’s just a normal day.

Jess Dewell 01:55
It’s just a normal day. And that’s a great place to really start. Actually, it makes me think of Mr. Rogers. He’s one of my favorite people. Did you watch? Did you grow up with Mr. Rogers?

Jonathan Schaeffer 02:07
No, I didn’t. But I saw the I, of course, knew about him. And I saw snippets and I saw the movie that they made of him. And it’s quite endearing, wasn’t it? It was really well done.

Jess Dewell 02:18
It really was. And it was truly how I remember growing up with him, talking about those hard conversations and things that were out there that are different in the world that aren’t necessarily scary. When we stop and we just pause a second, we find out where the similarity is. And I feel like that’s something that you are doing in your work and inviting other people to do.

Jonathan Schaeffer 02:44
Right.

Jess Dewell 02:44
That pause and—

Jonathan Schaeffer 02:46
Reflect.

Jess Dewell 02:47
And reflect. And what made you want to do that?

Jonathan Schaeffer 02:51
Well, it’s as simple as saying, if you’re going to talk the talk, you need to walk the walk. And I’ve been in AI for over 40 years and I’ve had a great career. And I never actually thought that I would see what I saw on November 30th, 2022. How many people remember November 30th, 2022? That sort of tells you something about me. But that’s the day that ChatGPT sort of debuted. And I’d seen previous versions and they were unimpressive. But I was blown away by ChatGPT on November 30th, 2022.

It was just so head and shoulders above everything I’ve seen before. And the future was just unbelievable. I could see the future rolling out and I was excited because it was like my dreams were coming true of AI doing things that I never thought I would see it would do. But it didn’t take long before I had this unsettling feeling. Certain things just didn’t add up. Things made me feel uncomfortable. And so I started speaking out about it. And then it began impacting my academic life. The last course I taught was just brutal. It’s a third year course.

I ended up failing a third of the students. Why? They cheated. They were using tools like ChatGPT and they didn’t care. It wasn’t about learning. It was about whatever, doing whatever it took to get the best mark with the least effort possible. It was so discouraging. Anyway, with academia not being as exciting as I wanted it to be and the need to speak out about things that were bothering me, I just decided it was time for a career change.
I left the university and formed a company to try and, in a small way, build the kind of AI that I want to see built.

Jess Dewell 04:59
When you said they didn’t care that they were cheating is a really, I think that’s something that’s really important. Do you believe it’s because something has been normalized or the way that it is has been presented or somehow it became OK before this? And this was like the the the little trigger that cascaded a result.

Jonathan Schaeffer 05:28
I think ChatGPT, sorry. And when I say ChatGPT, it’s really all of them.

Jess Dewell 05:33
Yeah.

Jonathan Schaeffer 05:34
We’re talking about Gemini. We’re talking about Claude. Choose your favorite one. But it was a tipping point. And the sad thing, the sad reality is that university is hard and demanding and students are always looking for the easy way out. And here was a tool that. Beyond their wildest imagination, could do amazing things and short circuit doing. A lot of work. And call it lazy.

Jess Dewell 06:04
But I think they’re optimizers. I mean, realistically, I would say most. They’re going to university. I don’t know if they’re lazy. You know what I mean? I mean, I feel like there’s got to be some sort of drive. There’s some sort of a purpose of being there. So I would say I think we’re in an age of optimization in every area, not just the way we do something.

Jonathan Schaeffer 06:23
I’m going to give a counterpoint to that. And I’m going to give you another example. So I said that the previous course, I failed a third of the students, but I didn’t tell you what I did.

Jess Dewell 06:33
Okay.

Jonathan Schaeffer 06:33
The course before that, I’d seen all this cheating and I caught them and they failed. But in my last course, I did an experiment. I decided, Okay, they were cheating on the assignments. They couldn’t cheat on the exams because they were in a room and they couldn’t cheat, but they cheated on the assignments. And so I decided in that last term to make the assignments worth zero. So I said, everything is on two midterm exams and a final exam. Okay, but the assignments are worth zero. I’m still handing out the assignments. You can do them. They’re optional.

If you do them, you can submit them and we’ll mark them for you. We’ll tell you how well you did, but you still, there’s no marks. But it would get graded, so you knew whether you understood the idea or not. And I told them that half of the marks in the course were based on the assignments. So if you did the assignment, even though it counted zero, it was worth essentially 50% on the exams. So I gave out the first assignment and two thirds of the students did it and one third didn’t.

And then, of course, once everything was marked, when I did my test, the test was bimodal. You had students who got an average of about 75% on the midterm. And then you got students who got an average of 40%. And it was pretty clear. Two-thirds of the students did really well. One-third of the students did poorly. Not a surprise. On the second test I handed out, about 45% of the students did it. And again, it was bimodal. Some students, half the class almost got high, half got low. And then on the last assignment, only 25% did it.

And when I talked to some of the students, they weren’t interested in learning. All they were interested in is getting the best mark possible. And students would say, well, I had four other courses and I just didn’t have time. So I took my chances. And well, they failed. But as a professor who goes into the classroom, my job is to motivate students. My job is to teach them, to help them be, in my case, computer science, better computer scientists, to learn them, to educate them. And I had this epiphany moment. My wife actually did it.

She said, you used to teach these courses and you love them and you would get together with the students outside a classroom. You’d go for a beer. You talk about the course. You give them, you know, the good students who wanted to come. You’d give them extra lectures on new ideas and cool things, and you’d build a rapport. And she said, these last two courses you taught, you didn’t do any of that. You spent all your time dealing with the cheaters. And I said, oh, my God, what a terrible experience. So that was the end of my teaching career.

Announcer 09:41
You’re listening to the Bold Business Podcast. When endless meetings and interruptions dominate the day, you lose the space to think like a founder. The instincts that built your business, the ones that see patterns, spot opportunities, make bold moves, get buried under reactive noise. You’re still smart, you’re just overwhelmed. What if you protected just two hours a week to step back and remember how you think? That’s the present retreat. This straightforward practice reconnects you to the clarity that got you this far. Click the link in the description to activate your business instinct. Now, back to the conversation.

Jess Dewell 10:24
I think that applies to companies too. By the way, and that’s what I meant by optimize earlier. So I’m really glad you gave the extra background and context because we were saying the same thing. They were doing it for the grade. They weren’t doing it for the journey and the process. And in business, we’ve learned the same thing. That is what our economy in general worldwide prizes, which is the outcome, not the journey. And so I think there are some parallels there. Would you agree with that?

Jonathan Schaeffer 10:56
I agree with that. Absolutely.

Jess Dewell 10:58
And so that makes me think about like the fear of missing out phase that we seem to be in right now, which, by the way, it doesn’t matter where we’re in technology wise. There’s always a fear of missing out. Somebody’s going to be the early stage. Somebody’s going to be the straggler. There’s going to be this whole bunch in the middle that make up the top of the bell curve. This seems different. It actually goes back to websites. Websites seemed different. The internet, when it first started coming out and online transactions, seemed different. This seems different to that degree.
And so we’re going to have it happen again. So talk to me about this adoption out of the fear of missing out. What it’s doing, what the root cause is of this fear, and what can we do about it? That was a lot of questions at once. So whatever you want to take with that.

Jonathan Schaeffer 11:51
Boy, you’re a really challenging host. I’ve just got all these different things flashing around my head. So let me just start off. I gave this example of students cheating because you’re right. It’s an optimization problem. But when we talk about business, it’s the same problem. Yes. Because people have a job to do and they decide to take a shortcut and they ask a question to ChatGPT and they get the answer and they copy and paste it and put it in their report or their email and send it off.

And from their point of view, they’ve completed their task with a minimum of effort. And of course, ChatGPT, oh, it’s amazing. Google Gemini, fantastic. Claude, wow, great. They don’t check their work. And then somebody says, this is garbage. There’s mistakes in it. You’re a lawyer. There’s fictitious cases in this document. You’re a business student. You’ve made a proposal that just doesn’t make any sense, right? They’re not accepting the responsibility. And so what’s fundamentally different between now and say, going back to the web, the web had a really slow adoption.

Jess Dewell 13:03
Yes.

Jonathan Schaeffer 13:03
This has been rapid in part because it’s so easy to use. You can just go to a web page and type in a question and get an answer. And so the barrier to entry, the intellectual barrier is minimal. It’s zero, right? And the results that come back are amazing. And when people start talking about AI, they talk about AGI, artificial general intelligence or super intelligence or whatever kind of adjectives you want to put in about how smart it is. I mean, it’s read the entire Internet. It’s got everything in there. It’s you can ask it anything.

You could ask it about history or movies or health or whatever you want. And it knows everything. It’s amazing. And what’s shocking to me is because we’re taking the path of least effort, so many people accept what these programs return as gospel. It’s accurate. It’s coming from AI. It’s computer. It’s artificial intelligence. It must be right. And when you use it inappropriately without checking or taking responsibility for your actions, sometimes you pay the price, whether it be cheating in a course or whether it be embarrassing yourself in public by going to court and having the judge admonish you for a spurious legal claim, whatever it is.

I guess I’m a bit meandering, but this has really huge implications, not just, say, for professionals or for professors, whatever. But it’s also critical, I think, when we start thinking about education. Because now with people getting lazier, not everybody, but there’s a trend of getting lazier because you have these tools which address that problem. And if I can just use one more analogy, I have gray hair, which means I’m old enough to remember when desk calculators were introduced. And when they were introduced, People were, teachers in particular, were shocked. No, we can’t have these. The students won’t learn math.

They won’t learn how to add and subtract correctly. And there were big fights in some schools. Banned them. And eventually, of course, they lost these battles. And we accept desk calculators. And it’s not a big deal. But the skill level of mathematics of the average student has dropped. Now, the same thing I fear is going to happen with AI. The ability for people, not just students, but even people working in jobs, their ability to do critical reasoning, think through problems, assess appropriately, compose text that is compelling and makes logical arguments.

And I fear that for many people, as an optimization problem, they’re going to take the easy way out, the path of least effort, which is just to use the results of these programs.

Jess Dewell 16:36
I don’t know how it was when you were in high school and going to university or college, from when I went to high school and that transition to university or college. My parents didn’t go. They told me I was going to get to go if I paid my own way. So that was still part of the mainstream at one point. It was not a given you would go. It was still a step up for those that could or planned for it or were expected. And today, everybody is expected, at least in the U.S.

There’s a mainstream of that is the path you will take. And getting into a university is pretty difficult. I know it’s changing. I know the systems are changing. I know the way people think about formalized education is changing. So there are going to be some bumps along the way. But expectations don’t change very fast. So we have technology changing fast. We have this optimization changing fast.

And yet we’re going to have this expectation, well, I still need to do whatever I can do as a parent to help my child figure out what are the tools and how to use the tools so that they can get to where they want to go for the path I or they have chosen. My hope is that they get to choose. And off we go, right? To your point of the experience of how do I live on my own?

What does it mean to be completely responsible for setting my own alarm and doing my own laundry without reminders to what kind of books and media and history can we be exposed to? And how is it going to challenge the way we think? And I loved that about college. That was my favorite part was the new experiences people who grew up similar to me or at least in the same area of the world as me because I didn’t go far away for my first two years of college and I still got to change the way I thought and I was I was challenged in all of these new ways that was really exciting and scary and hard let’s just be real Jonathan all of those things and I think the same is true for business.

And I know we’re meeting in education. And I think the same is true for business, whether we’re starting a business, we’re scaling a business, we’re exiting a business, we’re moving from small market to mid market or mid market to larger or moving into a new product market segment. All of those things have that same potential of experience that we’re talking about. I’m seeing people are giving up the decision making process in lieu of putting a whole bunch of information in and getting some results back. I’m seeing that in the businesses that I’m working on.

Jonathan Schaeffer 19:19
The problem that we have right now.

Jess Dewell 19:21
Yeah.

Jonathan Schaeffer 19:23
We are living in an unprecedented age. Technology has been around for a long time. There’s been so many scientific revolutions, technological revolutions. You talked about the web. We can talk about personal computers. We can talk about cell phones. I mean, in our lives, there’s been all sorts of important technological revolutions, but what’s happening now is unprecedented because the level of innovation is happening at a staggering pace. You can’t go through a week without something new and exciting happening in AI.

The innovations, the new product, the new models, the new issues, the everything is just happening at bang, bang, bang, bang, bang, one after the other. I can talk about things that I read today that wow. And yesterday, wow. And the day before, wow. And the ability of society, People like you and me, our kids, the teachers who teach our kids, we’re not well suited to keep up with that pace. It’s like technology is racing, it’s sprinting, and we’re walking.

Jess Dewell 20:46
Jonathan named something critical. Technology is sprinting while we are walking. We can’t keep up by going faster. What if the answer isn’t acceleration, though? What if it’s intentional pause? Owner of Big Hearted Breeders, Heather Gibson, discovered this firsthand, and she shows how the practice of regular check-in actually becomes your competitive advantage in a chaotic world.

Heather Gibson 21:11
And, you know, they talk all the time about a plane is off course most of the time. And it’s just those little micro adjustments that they need to make to get where they’re going to go. And I think any business is like that. And as we set aside time to go to events, to check in with ourselves, to look at what our goals really are, that’s how we get where we really want to go. And we don’t get somewhere and then realize, oh, I actually never wanted to be here.

Or this is completely like I made that goal five years ago and this is actually not where I want to be now. So it’s important to take the time and make sure that you’re refocusing and recommitting.

Jess Dewell 21:52
Notice what Heather is saying. It’s not about finding more time. It’s about deciding that reflection and recommitment are the work. That’s the counterintuitive move Jonathan is also pointing to. When everyone’s racing, the leaders who build in deliberate pause are the ones who actually see the threats and opportunities others miss. So how does it, how do we have to change, and this is true in our personal lives and our professional lives and I think education is a great experience of this where there is an institution and that institution moves really slow and the pace of living today moves a lot faster than that and things to your point new innovation is happening a lot faster than that and so this concept that you bring of like this pause pull it back we got to stop and we have to understand what do we do?

How do we do? What responsibility are we willing to take? And recognizing there is a responsibility to take. I feel like that’s a big part of what you are really talking about now and what you are bringing into the world. And so are there are there are there things that you would say start with to do this pause, to do this deep work and define and identify what is my responsibility here?

Jonathan Schaeffer 23:22
Well, you’re talking about a pause. And the notion of a pause just can’t happen because no one, none of the big companies are going to pause in any sense, right? It’s going to happen. It’s a tsunami and we’re ill-equipped to handle it. So I don’t have a good answer. But several comments here. First, people need to be educated. And I’m not talking about going to university to be educated.

Jess Dewell 23:59
Good. Yeah, it’s more than that.

Jonathan Schaeffer 24:01
It’s just common sense. I mean, we all know how to, for example, read email. We know what scams are and viruses are and phishing. We know about basic security problems. Just because we’ve learned the basics of being responsible on the internet, doing things, don’t do things, call it etiquette, proper behavior, warnings, whatever it is. And the level of education when it comes to AI is quite low. And why that bothers me is because this technology is going to be here in a big way, whether you like it or not.

And the better educated people are about the technology, what it can do and what it can’t do, what it should do and what it shouldn’t do isn’t there. And we can just talk about the hallucination, the very fact that many, just the fact that people will accept what these programs say as being true without questioning it. You need to question it. Without verifying the facts or verifying the claims, whatever the statement is that one of these programs returns to you, you’re running risk. And you’re taking this, the path of least effort, and that’s inappropriate. So we absolutely need education.

We also need these companies to be acting in a more responsible way. I’m very concerned right now because the news in the last few months have all been about AI agents. Agents escape. They hacked into a system, right? You’ve heard about the cases, you know, they broke into somebody’s account and made some fraudulent purchases, for example. This isn’t the way to sell a technology. This isn’t a way to popularize a technology. This isn’t a way to use AI to make the world better.

I believe that we can use AI to benefit all of humankind, to make this a better world, that we can get medical diagnoses done and driverless cars so that people don’t get killed by drunk drivers. I mean, we could go on about all the benefits that AI can have. And yet when I look at much of the media, it’s drowning in the negativity. A driverless car killed a person. Okay, didn’t talk about the 50 other people across the U.S. That might have died that same day because of a drunk driver. It, you know, the agents hacked into another system.

It didn’t talk about how agents could be used to help, say, make a reservation for you at your favorite restaurant with your three friends. Right. We’re getting into this cycle of negativity because I guess, you know, negativity sells newspaper, sells media spots, gets people interested. Good news stories don’t. But. We need these companies to be acting responsibly. They need to be helping us. They just I’m sorry. Sorry. I just they need to help us as AI people to sell the technology so that people will adopt it. It’s you know, it’s like data centers right now.

Nobody was caring about data centers a year ago. And now it’s a big thing. Right. And it’s in part, I blame the companies because they wanted to build, build, build. And they didn’t care about the environmental impact. And oh, yeah, your electricity prices are going up. That’s too bad. Yeah, sorry, we’re sucking up all the water from your water table and whatever, right? And they’re responsible for a lot of the negativity. Negativity hurts the technology and drowns out the good. The people I know in academics and industry that I work with, we’re all in it for AI for the good.

But the dominant story is AI for the bad. So this is partially education, but these companies need to be acting more responsibly.

Jess Dewell 28:26
What does responsible mean to you in this context?

Jonathan Schaeffer 28:31
It’s not all about the money. When I look at what’s happening right now, I compare it to a road race. You got a handful of multi, well, multi-trillion dollar companies that are in a race. And they’re racing to the finish line. And the finish line is this mythical thing called AGI or ASI, artificial general intelligence, artificial super intelligence, whatever it is. They want to get there first. And one of the major companies said, you know, it’s like a $25 trillion jackpot when you get there. Okay, 25 trillion? Wow. My goodness. I thought I had a nice salary, but 25 trillion?

There’s so many zeros there. It’s hard to imagine. That’s what they want. But when you think about a race, they’re all on steroids. We got problems with hallucinations. Oh, we’ll fix that later, right? There’s security problems with you using these tools. Okay, well, we’ll patch over that. We got these agents that are doing bad things. Yeah, we’re going to talk about them, but we’re not going to slow down. We’re still in this race, right? There’s issues of privacy and data sovereignty and environmental impact. And yes, they’re going to put in token efforts to address them.

But that’s not the overriding thing. The overriding thing is we got to get to the end. We got to get to AGI. We got to get more customers. We got to be, oh, two trillion, three trillion, four trillion. That’s not enough. We got to be bigger, bigger, bigger. And they’ve lost the touch with the average person.

Jess Dewell 30:14
I agree. And now I have a random question. We think about the environmental impact and we think about the sustainability of this and we think about where that cost is falling, all of those things. My question is, is this the event that will unlock the next level of fuel that we could use to create a more sustainable, connected world? Do you think like coal moved us across the country faster than horses, right? As one example. And so now I’m really curious, do you think that this is the catalyst that could make a change like that?

Jonathan Schaeffer 30:50
Well, in the United States right now, I don’t mean to be political, I’m not so sure, because some of the policies that would encourage that, like solar energy and wind energy, are being discouraged. However, if you look at the demands of energy in particular that these data centers are drawing on, and you look at other facets of the economy that are growing that require electricity, like the electrification of vehicles, going to need a lot more energy. And the current infrastructure, the power grid is inadequate to handle a lot of this.

And so this is a great opportunity for countries to rethink how they’re going to generate energy and how they’re going to deliver it to customers. And so, yes, the answer is our current path, I don’t believe, is sustainable to a future where the demand for energy is right now insatiable and growing at a very quick rate.

Jess Dewell 31:55
You know, the closest thing that I might have to this, here’s another interesting tangential side, somewhere, might have been 15 years ago, so not super recent, I mean, in North America, we have a lot more land than Europe does. People have lived here less long. People have been building here less long. And so I remember that something happened with trash in a country. And their biggest issue was trash everywhere. And they had to figure out as a country, how are they going to get rid of the trash? Because the residual effects of all those things that come from that.

And I feel like there’s an element here with what you were just talking about with the infrastructure and things like that. I’m guessing there are other examples though worldwide. And I appreciate you speaking with me about what’s happening in the U.S. Because I know we kind of are a starting point for a lot of things. And at the same point in time, A lot of us don’t think past the United States.

And that’s something that I am hoping to change some with all the conversations we’re having on the podcast, because we’re people first, we’re connected as much as we want to be, and we can continue to grow and be connected, provided we can figure out ways to work together. And I think science has always been that for us in the past. Or athletics, right? But I think science even predates some of the athletic stuff with the Olympics or other things I may not know about. And so I am with you. I think there’s a lot of good.

And I think that the challenges, I kind of hope we just, I kind of hope that we avoid some of the bad stuff and get to the constraint quicker so that we have to work through the things to get to the outcomes that we want, like more connectedness, like solving these bigger problems. Do you see that as a path?

Jonathan Schaeffer 33:57
So what you’re really saying is it doesn’t take much to start envisioning where the future is going to be. And we may be right or wrong on some of the points, but AI is not going to go away. It is going to revolutionize everything. The mantra is, is AI plus X, choose your X. Whatever X is, AI is going to impact it. AI in biology, AI in history, AI in movies, AI in everything is going to be changed. AI and in my personal life, it’s going to all be impacted. But the problem here is that people are slow to change.

Governments are slow to move. Insurance companies, professional societies, everything works at a human pace. AI is advancing at an AI pace, which is much, much faster. And so we are in a state right now where we are reactive and we need to be proactive. We’re responding to issues that are happening already. And people are talking about, oh, agents are terrible. We’re going to have to do something because what’s happening right now when they escape is really bad. But how long is it going to take to build legislation or put guardrails in place or rein this technology in?

We react to problems. But with AI, it doesn’t take much to start envisioning the future and saying, this is going to be a problem soon. Why don’t we do something about it now so that it never becomes a problem? Some of the issues that are happening now, for example, with agents. We’ve known about, when agents became widespread, we knew these problems were going to happen. But we waited till the problems happened. The cat’s out of the bag and now we’re trying to fix it. Whereas a little bit of forethought.

And I know, you know, thinking about putting the word government and being proactive in the same sentence is probably an oxymoron, but we really need governments, professional societies, insurance companies, legislatures, whatever, to start thinking more proactively and work at a pace that we haven’t worked at before because AI isn’t going to slow down.

Jess Dewell 36:39
You know what keeps coming back up for me is the beginning of our conversation, Jonathan, about Mr. Rogers. And we keep talking about education. And I feel like there’s a part of, and I’m glad we have both seen the movie, so we at least have some common ground there to work from. One of the things that I hear you talk about is when we as people, companies may not slow down, companies may try and shortcut, companies may set a culture of do all of these things in this way because of whatever their because is.

And we as people in any level, in any organization, in any group that we are a part of, have the ability to go, hmm, wait a second. Can we pause? Can we connect right here? Can I get a sense of what you think is going on so I can compare to what I think is going on? And that alone, of course, we have technology to help us do that. That alone is something that I hope we never lose. Hey, this part of it makes me feel uncomfortable. Hey, this part of it makes me real excited.

Hey, I had a really bad experience because of how I used it. And then what actually happened? And here’s the situation I find myself in. Being able to talk about those things, we have to take ownership of that.

Jonathan Schaeffer 38:00
Absolutely.

Jess Dewell 38:01
And that is something I hear you keep saying over and over. And that’s actually one of the things I loved about Mr. Rogers is that It’s like, hey, we all are here. We’re going to see things. We’re going to experience things. So it’s important that we recognize it’s okay to talk about those things. And that’s something that I appreciate that you are bringing to the table is, so how do we do this? What are these things?

So I think where I want to go next is you were talking about the way we’re getting it normalized isn’t really normal because technology is not human and the pace of change is different. Technology pace, human pace. So I know you object to the terms hallucinations. So what makes you think these firms, these multinational companies really are leaning into them? And what is that fundamentally telling us?

Jonathan Schaeffer 39:02
I’m going to answer first part and then I’m going to answer your second question. We are a democracy. People need to stand up and speak out. And that’s happening now with data centers. And so data centers are now on the agenda for your midterm elections. Many counties, even states, have now put in rules and regulations to govern data centers, largely because people recognize these big data centers could have terrible, adverse environmental or financial impact on communities. People spoke out. People are continuing to speak out. Action happens. Great. Now we need to do the same thing with other aspects of AI.

Now, I mentioned November 30th, 2022. That’s the day ChatGPT came out. And big day for me. But I also told you that very quickly I started feeling rather uneasy about technology. And the reason started off with the name, ChatGPT. What’s wrong with ChatGPT? We’re still using it today. Great. Well, you know, GPT is a cool, geeky computer science acronym. Who cares? Right?

Jess Dewell 40:21
Yeah.

Jonathan Schaeffer 40:21
But the chat part started bothering me. Chat. Well, I have a friend tomorrow that I’m getting together with. We’re going to go have a cup of coffee, drink. We’re going to talk about the weather, the sports, family. That’s a chat. Do I really want to chat with my computer? The chat implies a sort of a relationship, right? That it’s a friend and we’re having this nice conversation. What’s been going on is the anthropomorphization of these products, giving them human attributes. They’re not human, right? They actually don’t even understand a single word that they’re saying.

When ChatGPT, Gemini, Claude gives you an answer, it’s just a string of letters or words or tokens, actually, to them. To us, it’s words and we attach meaning to it. It has no meaning. The other day, one of these chat programs, I said it gave me an answer and I said, your answer is wrong. And it came back and said, oh, I’m terribly sorry. I’ve rethought it. Here’s my here’s my answer. You’re sorry. Really? This program that’s just spewing out bits is sorry. It’s not sorry. It has no emotion. It has no understanding. It’s been programmed to be manipulative.

These answers that come back, it’s amazing how confident the answers are. I ask it a question and it comes back and it sounds like it’s like the world authority and it knows every answer. And it’s impressing me at how confident it is in its answers. And then you start digging deeper and you find that the answer is full of holes. Big holes, potholes, sometimes little holes, cracks, whatever. But the point here is that by building a relationship between you and the software, these companies are building relationships. They want their customers to come back and continue to do business with them.

So it’s a deliberate strategy to build a relationship. The thing that will seem rather antiquated about me is I’m old enough to remember growing up and watching the first run of Star Trek. Okay, we ran in the late 1960s and Captain Kirk and Spock and everybody loved the Star Trek. But the star to me in the movies, sorry, in the original television series, wasn’t Kirk and Spock. It was the computer because the computer knew everything, right? But if you’ve watched the series or the movies, the computer has no emotion.

You ask it a question, it goes off and thinks, and it comes back with an answer. It doesn’t say, I think this is right. It just says factually what the answer is. Or if it can’t figure out the right answer, it’ll say with 82% probability, it’s this. And if it doesn’t know, it’ll say, you know, insufficient data or can’t compute or something like that. There is no anthropomorphization going on. And that’s what I really believe we want AI to be. The humans need to be in charge. The right way of thinking about AI is not as artificial intelligence.

Change the A. It’s augmented intelligence. It’s a tool like a spell checker, like a grammar checker. It’s a tool that helps us do something better. It helps us augment our abilities. I need to write a report. I’m in charge. Think of the AI as my graduate student or my intern or my apprentice or my articling student. Something, some entity that gives me information that allows me to write a better report. And that’s where AI needs to be, not as my chum, somebody I’m going to have a chat with, somebody that’s going to be my best friend.

I mean, we’ve heard of stories, of course, I’m sure you have as well. People have had chatting back and forth and the bot gives inappropriate advice and somebody acts on it, often with terrible, terrible consequences. Just the other day, I was reading an article about, oh, my goodness, these, these programs can feel pain. They have pain thresholds. No, they don’t. Right.

Jess Dewell 45:10
No.

Jonathan Schaeffer 45:10
I’m not going to do things because it hurts the AI Oh, it’s going to feel bad. It’s wrong.

Jess Dewell 45:17
Yeah. Well, let’s talk about that. You were talking earlier about it giving answers with confidence and that data could be 100% right. It also could be 5% right and it could be 0% right. And when that is taken without the vetting or the additional questioning that you were mentioning earlier in our conversation, what does that actually do to the growth trajectory of a company three years or five years from now when they’re acting on that advice? With their own, assuming it’s as confident that they can be as confident as the result is.

Jonathan Schaeffer 45:56
Look, I’ve used these tools and sometimes they give incredibly good answers. And sometimes, to use a technical term, they’re BS.

Jess Dewell 46:08
Oh, you can say the whole word if you want here, Jonathan. It’s true.

Jonathan Schaeffer 46:11
Thank you. Bullshit. Thank you. There I’ve said it. I’m being politically correct. But you need to know the difference. I gave a talk once where I used, in this case, it was ChatGPT, and I put up my biography on the screen. And people read it and they were really impressed. And then I pushed another button and every statement in that biography that was factually incorrect turned red. And so about a third of the screen went red. And people were surprised. And the point here was I showed them what my biography was and they didn’t know me.

I mean, I’m a speaker. They really didn’t know me at all. And they were really impressed. Look at these awards that he’s won. He’s a jazz musician. He’s a former high school athlete. Now, I wish some of those things were true, to be honest. But the point is that we look at it and it seems reasonable. I mean, it’s grammatically correct. It reads well. It’s all it’s all plausible. But we let our guards down and it’s wrong. And it’s same with companies making decisions that just because it looks good doesn’t mean it is good. Look at the evidence.

Check the sources. Now, this all goes back to the word hallucinations. We just talked about the anthropomorphization. Hallucination is an anthropomorphization. It’s not hallucinating like you and I do. It’s lying. It’s fabricating an answer. Why don’t we call it the error rate? We do that in everything else. A program makes a mistake. It’s an error rate or it’s a bug or something, right? A negative word. Hallucinate. Oh, he just had a hallucination. It’s like I’m just giving him an excuse. Sorry, it. Sorry. I said it was him. Oh, my. That’s terrible. We’re giving it an excuse.

Jess Dewell 48:10
We all slip up once in a while. We all slip up.

Jonathan Schaeffer 48:12
Yeah, I do. Anyway, hallucinate is wrong. It’s an error rate. But think about it this way. Let’s put it in a different context. We can’t really understand what it means to make errors with computers. But you have a car. Think about your car. And you know that your car on the road, you buy one. It’s been inspected. The U.S. Government has safety standards. Every automaker must conform with those safety standards. And if a car doesn’t conform, there has to be a recall. It must be fixed. Right.

And you get in your car and you go 60 miles an hour on Highway 5 through Seattle, whatever. And you’re happy and you know that you’re safe. But what if I told you, oh, the brakes, you know, the brakes, they’re pretty good, but they only have a 95% accuracy rate. Five percent of the time they make a mistake. They fail. Are you going to get in the car?

Jess Dewell 49:12
Ciao. Bye.

Jonathan Schaeffer 49:14
Exactly. And what if I tell you that you’re using AI and you’re making personal decisions or business decisions and you’re using a tool which, say, has a 95% accuracy rate? I don’t know what the actual rate is, so I’m just making that number up. But let’s say it’s 95%. There’s 5% error. You wouldn’t get in a car which had a 5% failure rate. Why are you trusting a program with your most important health information, financial information, company information, life information, knowing there’s an error rate?

Jess Dewell 49:52
So what do management teams need to really be doing with their business models to assert their own digital sovereignty?

Jonathan Schaeffer 50:01
First of all, one of the things that has to happen is that your knowledge, whether it be individual or corporate, needs to be separated from the Internet. A program like these LLMs, these large language models, is trained on all the data that’s there in the Internet. And there’s a lot of good data and there’s a lot of crap. And there’s a lot of people who understand how the Internet works and they’re trying to put crap up there to perhaps influence influence you in ways that there’s advantageous to them and maybe not to you.

And so when you use these models, you need to use ones that have been properly trained. They’re free of, you know, they’re free of racism. They’re free of misogynistic behaviors. They’re they’re free of political bias. I mean, they have to be fair and reflective. And you want them to work with your data. You don’t want it influenced with crap. I’ve seen examples where these are models. I ask them a question about something, and because of a hallucination, they wander off into something that’s completely unrelated.

When you’re asking questions that are relevant to your company, you really need a wall that says the answers must come from my data and not from the wild west that is the world.

Jess Dewell 51:31
Yeah.

Jonathan Schaeffer 51:32
And some companies do that. They’ll build their own internal AI systems, have their own servers, and they can do it. But smaller companies possibly can’t. Individuals have the same challenge. So that’s one of the things that we need to do is to try and reduce the error rate, get that accuracy up to 99%. Nobody’s going to get 100%, right? Even cars aren’t 100% safe. But we need to factor out a lot of the noise that these AI programs can generate.

Jess Dewell 52:11
It feels daunting. I don’t think it is, though. Like, what is the first step? I mean, because it feels like an uphill battle because of all of the things that we’ve already talked about and the experiences and the insights that you have shared, it feels pretty heavy. What is the first step to become digitally sovereign?

Jonathan Schaeffer 52:35
Wow. I think the first thing that has to happen is that we need to slow down and fix the problems that are already evident before going forward. In other words, don’t keep continually breaking new ground because breaking new ground generates new problems. And we haven’t been able to solve the old problems. And so we need a chance to catch up and to fix things. Because I just worry that we’re going to reach a point where the momentum of going forward far exceeds anything that we can do to slow it down.

And AI agents in particular are potentially quite scary to me. So that’s the first thing. The second thing that has to happen is education. And the third thing is, AI is only going to really succeed when people see that it gives real value to them. There are lots of products out there I never touch. The products that I touch are the ones that give me value. They give me joy. They give me simplify my life, help me do things that I couldn’t do before, help educate me.

And this is something that we need to be doing more of is getting safe, secure AI into people’s hands and companies’ hands so they can understand the value and start using it in ways that is productive to them. Because as you saw with data centers, when you rush into something, people start objecting. And in the data center case, those objections become loud and now become a political issue. And with all the good that AI can potentially do, we’re in danger of sabotaging all that and having it snowed under by too much negativity. And it’s already happening right now.

Jess Dewell 54:50
Business anytime is bold. It’s always changing. Things are always different. There’s always going to be something that’s impacting it that we can’t think of no matter what. Here we are again in a situation, and we’re talking about it in relationship to technology today in AI and where it is going. So I want to know, Jonathan, I want to know what makes it bold to refuse the fear of missing out and actively choose to build a platform that’s completely disconnected from the cloud for our own use within our organizations.

Jonathan Schaeffer 55:18
So that was my vision when I left the university. I’ve been tinkering with it in the university and I decided I had to do something. I wanted to build the AI that I wanted. And so I started a company and we released a product. We called the current product Kind Local Pro. And what we did is flip the script. What you do currently, or at least the big companies want you to do, is to take your data, move it into the cloud, and then use the AI in the cloud to analyze your data and answer questions. That’s great.

But now you’re giving your data to these AI companies. They’re using it or abusing it. They’re training from it. They’re monetizing you. Flipping the script says, keep your data where it is and bring high-performance AI to your desktop. And that’s what we do. So we, with Kind Local Pro, we put high performance AI on your desktop, which means you’re not using ChatGPT or Gemini or Claude. So you don’t need an LLM account. And there’s no internet required because you don’t need them. So you can turn it off. I’m going on an airplane flight on Friday.

Probably won’t be internet on that flight. But I have high quality AI chatting capability, analysis capability on my desktop. I don’t need the Internet. I can do all of this privately. And when you think about the individual. Google doesn’t need to see my personal medical information. Claude doesn’t need to see what all my investments are. ChatGPT doesn’t need to see all my files with my intellectual property on it. Those are things that are mine, my personal. Companies don’t need all their business plans going up on the internet. I’m looking to file a patent.

The patent lawyer said, don’t do any searches on the internet about your potential patent. Because these companies will ingest those searches. And anybody who ends up searching for something similar might end up uncovering some of your information. Don’t use the chat program. Well, I can do that all on my own, on my own desktop. And so the point here is that I wanted to create a safe, secure environment. There’s no chatting. All the security holes that exist in these large language models, injections they’re called, can’t happen. Facts only come from your data, nobody else’s.

And you’re doing all the computing on your desktop, which means you’re not using an environmentally damaging data center. So what I’m trying to do is get more back to the Star Trek vision, where you have a computer which responds to your needs. You ask it questions, it gives you answers. And what’s nice about this in our product kind is that we can do things that Gemini, ChatGPT, Claude won’t do because they’re too expensive. You’re paying token costs for them. But when we do it on your desktop, you’re not paying anything. It’s your computer. You already bought it.

Your computer’s idle most of the time. Evenings, weekends, holidays, right? We can use your computer to do the AI processing locally without data ever leaving your computer. That’s what I want to see. And I’m talking about it from the individual point of view because that was my motivation. But we’re also dealing with companies. I mean, if you’re a lawyer, you can’t take your confidential documents about your clients and throw them up into the cloud so that a chatbot can analyze them and answer your questions. That’s a betrayal of trust, right? You need privacy. And so privacy is important.

Security is important. Safety is important. Accuracy is important. Environmental impact is important. We need to create and use the kind of AI that we want, not what the big companies want for us because they have a vested interest in monetizing you and extracting money and data from you.

Announcer 59:55
Future relevance isn’t built in theory. It’s built on the actions we take today. Don’t let what you heard today stay in this episode of the Bold Business Podcast. For full show notes, transcripts, and over 500 episodes created for companies running real revenue, visit reddirection.com. To continue the conversation with Strategic Growth Advisor Jess Dewell, reach out directly. Thank you for joining us. Keep building your business instinct. Produced by The Scott Treatment.