The Marketing AI Show is back! The smart CRM market is evolving…and so are marketers and businesses with the help of AI.
First comes ChatSpot, then comes Salesforce Einstein GPT
Coming on the heels of HubSpot’s ChatSpot announcement, Salesforce just announced Einstein GPT, a generative AI tool for its market-leading CRM. The tool, which is currently in closed pilot, creates content across marketing, sales, and service use cases.
Salesforce’s communications say, “Einstein GPT will infuse Salesforce’s proprietary AI models with generative AI technology from an ecosystem of partners and real-time data from the Salesforce Data Cloud, which ingests, harmonizes, and unifies all of a company’s customer data.”
They say Einstein GPT can generate personalized emails, generate specific responses for customer service teams, generate targeted content, and auto-generate code for developers. In the same breath, the company also announced a $250 million Generative AI fund through its venture arm.
The value (or lack thereof) gained by AI is dependent on three factors.
Paul recently published a post on an AI topic framing his idea of “the law of uneven AI distribution.” In it, he wrote: “The Law: The value you gain from AI, and how quickly and consistently that value is realized, is directly proportional to your understanding of, access to, and acceptance of the technology.”
This uneven distribution will create dramatic differences in people’s experiences with and perceptions of AI. And it’s all dependent on three factors: how well you understand AI, the level of access you have to AI, and how much you accept the radical changes that AI will bring about in business and society.
Do we need to fill the time saved by AI with more…work?
When we talk about AI, we often hear that the wondrous productivity gains produced by AI technology will give us back more time, in turn making our lives less busy and more fulfilling.
And these productivity gains are valuable. Venture fund ARK Invest predicts that we could boost the productivity of the average knowledge worker by 140% with AI, which would create $56 trillion in value globally. But a new article from the Centre for International Governance Innovation challenges the idea that AI will liberate our time and goes so far as to call the AI productivity narrative “a lie.”
However, history has shown that efficiencies often heighten expectations and standards. How can we as marketers, business leaders, and humans, ensure we aren’t exacerbating Parkinson’s law by adding to the idea that “work expands so as to fill the time available for its completion?” How can we invest that time in things we want to do? This is a topic we all need to listen to!
Listen to this week’s episode on your favorite podcast player, and be sure to explore the links below for more thoughts and perspectives on these important topics.
00:03:43 — Salesforce Einstein GPT
00:13:20 — The Law of Uneven AI Distribution
00:23:54 — Why the productivity narrative is a lie
00:34:59 — Silicon Valley Bank
00:39:39 — The dangers of putting ChatGPT into everything
Links referenced in the show
- Salesforce Einstein GPT and the Smart CRM Market
- The Law of Uneven AI Distribution
- Why the AI Productivity Narrative Is a Lie
- The Dangers of the Race to Put ChatGPT Into Everything
Watch the Video
Read the Interview Transcription
Disclaimer: This transcription was written by AI, thanks to Descript, and has not been edited for content.
[00:00:00] Paul Roetzer: We have to have these conversations that AI isn’t just another tool to get more out of every hour of every workday. It’s we can achieve more productive professional lives. We can create more value in our professional lives, but in an ideal world, we get to redistribute some of that time back to ourselves for our own health and wellness and just enjoyment because the time is limited.
[00:00:26] Paul Roetzer: Welcome to the Marketing AI Show, the podcast that helps your business grow smarter by making artificial intelligence approachable and actionable. You’ll hear from top authors, entrepreneurs, researchers, and executives as they share case studies, strategies, and technologies that have the power to transform your business and your career.
[00:00:46] Paul Roetzer: My name is Paul Roetzer. I’m the founder of Marketing AI Institute, and I’m your host.
[00:00:55] Paul Roetzer: Welcome to episode 38 of the Marketing AI Show.
[00:00:58] Paul Roetzer: I’m your host, Paul Roetzer, along with my co-host Mike Kaput. Good morning, Michael. How are you today? Nice. Doing well,
[00:01:05] Mike Kaput: Paul. Uh, there’s been a lot going on, so I feel like my head’s already spinning and it’s 10:00 AM on
[00:01:10] Paul Roetzer: Monday. . Luckily, we are not dealing. Worst fallout from the Silicon Valley Bank thing, which we will talk about maybe at the end here.
[00:01:19] Paul Roetzer: But, uh, it could have been a crazier start to the morning for everyone. , we’re, we’re obviously recording this on, uh, March 13th, Monday, March 13th. Uh, you know, it comes out, uh, on Tuesdays, but, uh, yeah, it was an interesting weekend for anyone in the VC world or in the startup world or anyone connected to Silicon Valley Bank.
[00:01:36] Paul Roetzer: So, uh, yeah, we’ll touch on that at the end. Yes, we are another week in AI and plenty to talk about before we get started. This episode is brought to you by our AI for Writer Summit. If you’re a regular listener to the podcast, you’ve heard us talk about this. It’s ai um, writer summit.com. The event is March 30th.
[00:01:56] Paul Roetzer: It is a free virtual event, uh, runs from 12 to four Eastern time. And the whole idea here is just to bring everybody together and try and figure out what is going on in the writing space. So as chat, G p T and these other language tools have emerged, um, these different AI writing tools and editing tools.
[00:02:15] Paul Roetzer: There’s lots of questions, uh, about the impact it’s gonna have on writers, on editors, on content teams, on media companies, on publishers, on brands and content strategies. And so we’re trying to bring people together and really push this conversation forward. We’ll be talking about different AI writing technologies, looking at use cases, uh, looking at the impact of writers and teams going through impact potentially on career paths and how to kind of infuse AI into those career path.
[00:02:43] Paul Roetzer: And also getting into some of the negative implications and some of the challenges and fears people have around it. So that event’s coming up March 30th. We originally thought we’d have 500 to a thousand. We’re up to, I think 2300 registered as of this morning. We had 600 in the last week registered. So it’s obviously something that’s on a lot of people’s minds.
[00:03:04] Paul Roetzer: Uh, again, there is a free registration option, thanks to writer, which is the presenting sponsor. So thank you to writer for making that possible. And then we also have, uh, a number of other sponsors we wanted to just recognize briefly go Charlie Vila, hyper Wright. Right? Raa I. Demand, well gloss AI and copy all people who stepped in as supporting sponsors for the event.
[00:03:27] Paul Roetzer: So we thank all of them, uh, for helping make that event possible for everybody as we’re all trying to kind of figure out what’s going on in this writing space. So if you’re new to the show, Mike and I pick three big topics in AI each week and uh, I’m gonna turn over to Mike and get us. Thanks, Paul.
[00:03:43] Mike Kaput: So first up we have a big announcement from Salesforce.
[00:03:48] Mike Kaput: Salesforce just announced what they’re calling Einstein, g P T, and that’s a generative AI tool for Salesforce’s market leading c r M. So this tool is currently in closed pilot, but what it does is it creates content across marketing, sales, and service use cases. Using the same types of generative AI we’ve seen in some of the leading market tools like Chat, G B T Dolly two, what have you.
[00:04:13] Mike Kaput: So Salesforce says that Einstein G B T will actually infuse Salesforce’s proprietary AI models with generative AI technology from an ecosystem of partners. And realtime data from the Salesforce data cloud and for example, Einstein, G P T could do something like generate personalized emails for salespeople to send to customers, generate specific responses for customer service professionals or more quickly answer customer questions and generate targeted content for marketers so they can increase their campaign response rates.
[00:04:50] Mike Kaput: And it can also even auto-generate code for develop. In the same breath as this announcement, the company also announced a 250 million generative AI fund through its venture arm. So obviously last week we talked about HubSpot’s Chats Spott, which is a similar tool for the HubSpot C R M. So this announcement is coming right on the heels of that as we see these big platform providers get into the generative AI game.
[00:05:20] Mike Kaput: So first up, Paul, I wanted to ask you. With all these announcements, are we seeing the birth of a smart C R M
[00:05:28] Paul Roetzer: category? Yeah. That would appear to definitely be the case and but anybody who’s followed us for a long time knows, what we always say is AI is just smarter technology. So the reality is every software category, there’s going to be a smart X category to it.
[00:05:44] Paul Roetzer: Like everything is gonna become smarter, everything’s gonna be infused with. So this seemed like a very obvious starting point for, um, these platforms. So again, as you mentioned, Dharmesh and HubSpot introduced, uh, chats Spott last Monday. So that was a topic of conversation and then this was backed up on Tuesday.
[00:06:04] Paul Roetzer: Um, couple things jumped out to me here. The generative AI fund in particular, I find interesting. So it’s a 250 million. They announced an integration kind of right out, out of the box with OpenAI, uh, for this Einstein g p t, but two of their first four investments were in other language model companies with anthro and cohere, which are two names you’ve, you’re probably familiar with if you’ve listened to the podcast.
[00:06:30] Paul Roetzer: So, You know, language is obviously the lead here when we’re talking about generative ai, it, it covers language, images, video, audio, and code would be the other one that maybe we haven’t historically thrown in there. But I, I think you talk about generative AI without talking about code. So, you know, they’re obviously making the major play right now into language because that was what happened, you know, last year, 2022 was sort of the year of language.
[00:06:57] Paul Roetzer: As well as image. But you know, I think we’re starting to see that infused video, I think in 2023 is gonna be, you know, on par with what we saw with image and language last year. So, yeah, it just, it seems like an obvious play. I’m curious to see how quickly these things roll out. Cuz even chat spott is still sort of in, you know, limited rollout.
[00:07:16] Paul Roetzer: I think Darash had said he was doing like a hundred people a day, but then I, I thought I saw a. Over the weekend that he was gonna, anyone on the wait list was gonna have access this week maybe. So it seems like chatbot’s really, you know, starting to roll out as well. Um, but I think this is gonna be rapid innovation.
[00:07:32] Paul Roetzer: We’re gonna see a lot of this stuff playing out this year, but it makes perfect sense. C r m marketing automation, um, that stuff’s gonna get smarter for
[00:07:39] Mike Kaput: sure. So obviously the AI capabilities both depend on kind of your own use cases and what you’re using the platform for, but at a really high level, could you maybe walk us through some examples of what kind of impact tools like Einstein G P T or Chats Bott would have on kind of my daily or weekly work?
[00:07:59] Mike Kaput: As marketer using a C R M system.
[00:08:02] Paul Roetzer: I think it’s to be determined until you get in there and see what these things are actually capable of. You know, there’s lots of promises being made, or at least presenting what the use cases could be. The question becomes do they actually fulfill that promise? Like does it actually do what it’s proposing to do?
[00:08:17] Paul Roetzer: So I think there’s gonna be a lot of experimentation to see if these tools are really as far along as the brand messaging makes them sound. My guess is they’re not, um, usually there’s a bit of a product development curve here that has to happen. These are gonna be, you know, a Chads spot to Dharma’s credit.
[00:08:32] Paul Roetzer: He’s positioning as an alpha product. So it’s like, don’t, you know, it’s not gonna do exactly what you think it’s gonna do all the time, but it’s all, it’s gonna be about rapid innovation. The same time, you know, I’ve talked with some larger enterprises in the last week or. In particular since the Einstein thing came out, but we’ve always told, especially bigger enterprises, is start with your core tech stack.
[00:08:52] Paul Roetzer: When you’re thinking about AI adoption and, and like scaling AI in your organization, you always start with the core companies that are already there. So if you use Salesforce or HubSpot or Oracle or whatever it is, whatever your core stack is, specifically your crm. You wanna start with those companies and talk to your reps and say, do you have smarter features we should be using?
[00:09:13] Paul Roetzer: Like, are, are there AI capabilities in the product that we aren’t using that can make us more efficient today without having to go and complicate our tech stack by going and finding five or 10 new tools to all these different use cases? So I think the biggest impact is gonna be the integration of these capabilities into your workflows with the existing technologies you.
[00:09:35] Paul Roetzer: Um, the other thing is you might get in there and realize, wow, I, I wish it worked the way they said it would work. Maybe since it’s not, I need to go find a tool that does do this. And so, you know, it really helps you start to visualize the way that AI can be infused into your daily workflow. And these products are gonna get better really fast.
[00:09:53] Paul Roetzer: Um, you know, we’re talking. These things are being built largely on two year old technology. So if you look at open AI and the integration of like chat G P T through their APIs into these products, that tech is a couple years old now. It’s improving and we’re getting, you know, incremental improvements to it from what open AI is releasing.
[00:10:13] Paul Roetzer: But the expectation is in the coming months, we’re gonna see some leaps forward in these capabilities and just gonna kind of keep resetting what these platforms are able to. So that’s
[00:10:24] Mike Kaput: kind of tackling what your marketing practitioner should be thinking about when it comes to a smarter crm. If I’m kind of a leader or an executive, Just now seeing these features come out, thinking about my database, my C R M system, and the third parties I may work with in terms of C R M support.
[00:10:45] Mike Kaput: I mean, that could be anyone from a developer in Salesforce, HubSpot agencies that specialize in these platforms and getting more value out of them. What should I be thinking about here? Do I need to be evolving
[00:10:57] Paul Roetzer: my approach? Yeah, I, I mean, It’s gonna be a really interesting time for director level and above to be trying to think through all this stuff because it does have an impact on your tech stack.
[00:11:09] Paul Roetzer: It has an impact on your own team and maybe the need to upskill or, you know, acquire some, some new talent. And it certainly has an impact on your partner ecosystem, specifically if you work with outside agencies and consultants. So, you know, right now it’s nice to have an agency or consultant that understands.
[00:11:30] Paul Roetzer: Hmm. Uh, I see that quickly becoming a need to have. And the reality is there just aren’t that many agencies and consultants that understand this stuff. So it, it’s gonna be a challenging environment as we go through this kind of very disruptive phase where we’re trying to understand this stuff. But like I said, it, it’s gonna become essential really fast.
[00:11:49] Paul Roetzer: There’s a lot of pressure on marketing and business leaders to figure this stuff out, right. And it’s at all levels. I mean, we’ve talked with lots of, uh, leaders at different size organizations. Everyone is trying to figure this out. Now, some of them think they’re trying to figure out chat. G P t like we’ve talked about before, that chat, G p t, is just a tool in, in this massive ecosystem of AI applications and capabilities.
[00:12:14] Paul Roetzer: And so people are kind of misguided in thinking that if they solve for AI writing, then that that’s it. They’ve solved for ai. So, yeah, I still think that there’s a massive gap in knowledge in the marketplace where these company and, you know, business leaders aren’t even sure what they’re solving for.
[00:12:32] Paul Roetzer: Like they don’t know what AI actually is. They don’t know what business problems it can really solve. In many cases, they just know it can now write content and they’re trying to like scramble to figure that part out. But, but that’s really just the entry point for a lot of people and. Yeah, I think leaders are gonna struggle a little bit.
[00:12:48] Paul Roetzer: That’s why, you know, at our marketing AI conference in July, we have an OP two optional workshops. Um, the one is applied ai, which is you’re a practitioner and you need to find use cases and tools right now to start testing. And the other is the strategic AI leader. And that’s the premise of that one that I’m actually gonna be leading, which is like, what do we do about this?
[00:13:08] Paul Roetzer: What’s the impact on operations, on workflow, on talent, on overall business strategy? And so I think that’s the stuff that’s critical in 2023 for people to be solving for.
[00:13:20] Mike Kaput: So next up, we have a AI topic that affects everyone regardless of profession or industry or level of AI understanding. And Paul, you had actually published a post and created the idea behind this topic and.
[00:13:36] Mike Kaput: It’s called the Law of Uneven AI Distribution, and what you wrote is this quote, the law, the value you gained from ai and how quickly and consistently that value is realized is directly proportional to your understanding of access to. And acceptance of the technology. And kinda what that means is that this uneven distribution of AI is going to create dramatic differences in people’s experiences with and perceptions about artificial intelligence.
[00:14:06] Mike Kaput: And it’s going to profoundly impact how much you can reap the rewards, the benefits of ai. In your personal life and in your professional life, how much value your company can extract from the technology and what your AI journey actually looks like. So like you alluded to, there’s these three factors that influence the distribution of AI’s benefits, how well you understand it.
[00:14:28] Mike Kaput: Firstly, Second, the access you have to ai. And third, how much you accept the radical changes that AI is going to bring about in business and society. So first up, can you kind of unpack this idea behind the law of uneven AI distribution for us a bit
[00:14:45] Paul Roetzer: more? It’s, it’s something I’ve been thinking about for a long time and I just wasn’t sure how to put it into words.
[00:14:50] Paul Roetzer: And then just like ongoing conversations sort of drove me to have to like figure this out and be able to explain it in a logical way. Um, and a couple of things that, that really drove. The creation of this was conversations I’ve been having with, uh, leaders at enterprises and highly regulated industries.
[00:15:11] Paul Roetzer: Uh, were there, like, we don’t have access to chat chip pt. No, I haven’t tried it. We don’t, we can’t use it. Um, it was shut off like on day two. So you realize like, oh, okay. There, there are issues with one, people aren’t even gonna be allowed to use these tools in their organization. And so they are inherently not going to receive the benefits of AI because they’re not allowed to use.
[00:15:33] Paul Roetzer: So that was top of mind. Then I was hearing examples of people getting it shut off because their organization realized they were putting confidential information into these tools to help train them or to get better results out of them. So, you know, copy and paste some confidential proprietary information, throw it into Chad G P T and get an output.
[00:15:53] Paul Roetzer: Not thinking about the fact that, oh wait, we don’t even know what chat G B T or OpenAI does with this data. Mm-hmm. So, again, You can see why organizations would take these steps. The same was happening in higher education and even in middle schools and primary education. So I’ve been having conversations with leaders of universities and then leaders of school districts who are trying to figure out what do we even do about this stuff?
[00:16:16] Paul Roetzer: Do we teach kids about it? Do we limit their access? Do we embrace it and actually help them understand it? And so again, those students ability to use these tools and understand those is actually being dictated in part. What was being allowed to be used. Then I looked at my own personal use. There’s all these fascinating AI tools.
[00:16:36] Paul Roetzer: There’s things I see every day, like the one I, I think you and I talked about, like the infusion of it into like Google Sheets, where it just starts like doing narratives based on your data. And my first reaction is, well, who is this company that I’m giving access to my data? So there’s all kinds of confidential information, whether it’s my personal life or or, or in business where I have that stuff sitting in Google Sheets or a Google.
[00:16:59] Paul Roetzer: And like, what company am I allowing to access that information and what are they doing with that data? Is it gonna then be extracted to train future models? Is like my data gonna end up in some training set? So because I know how these models learn, I, I realize I might be a bit of an outlier of like my unwillingness to give access to these things in my personal.
[00:17:20] Paul Roetzer: Like lens, for example, when that was blown up, I wouldn’t use it. Like I, I didn’t feel comfortable with giving my data to this company that I didn’t really know. And so I just started thinking about these different examples of why I might not receive the benefits of AI or even get to the point where I even comprehend it cuz I can’t use it.
[00:17:41] Paul Roetzer: And that was really like the impetus for why I kind of like started thinking like, how do I explain this in words and kind of put it in a digestible way. That outlines for people like the things that are gonna affect your ability to benefit from ai.
[00:17:57] Mike Kaput: That makes a ton of sense. And so just to kind of bring this home and connect the dots here for people it’s like, We’re, the reason we’re talking about this is not only because these issues are extremely important when you are adopting ai, but really the high level idea here it sounds like is that new future will be determined by how AI is adopted and used both in our personal lives and our business lives.
[00:18:19] Mike Kaput: So, It sounds like that’s really why this matters. Like obviously everyone’s really busy trying to figure out all the different tools. Some, some don’t even have time to sit around and worry about ai. But why, why is this important for marketers and just individuals to really address right now?
[00:18:37] Paul Roetzer: Well, you know, I think the promise, when we look at these major research labs like OpenAI and DeepMind and, you know, some of these other stability, ai, you know, where they’re, they’re all pursuing.
[00:18:48] Paul Roetzer: Intelligence, like these smarter applications under the premise of like abundance, like that if we solve for intelligence and we build all these smarter applications, we’re just gonna live better, more fulfilled lives where we just have limit, like limitless access to, to goods and services and wealth and you know, it’s kind of this utopian view of where AI.
[00:19:10] Paul Roetzer: And so that was the other thing I started thinking about. I was like, well, that’s not gonna be true. Like the rich may get richer because they understand it, they have access to it, and they’re generally accepting of what it does for them. So yes, like organizations that are very forward thinking, those people will benefit, they’ll build smarter companies.
[00:19:29] Paul Roetzer: Mm-hmm. individuals, you know who, people who are at kind of the forefront of. They too will benefit. But then I just started thinking about the distribution across the rest of society and the rest of the business world. It’s like, well, there’s a very decent chance this promise of abundance is going to be unevenly distributed to the people who already have abundance in their lives.
[00:19:48] Paul Roetzer: And so I think that, you know, the key for me was we need to create an awareness that this isn’t an even playing field, that AI isn’t going to be. And now in some cases, you will willingly choose not to take those benefits. So like for example, we talk. in your personal life. The benefits, personalization, convenience, abundance, fulfillment, happiness.
[00:20:09] Paul Roetzer: If you allow these AI tools to create all the things that they’re capable of creating for you, if you give it access to your life, you understand what it is, and you’re willing to accept that you’re giving some data up in exchange for this and maybe some of your privacy, but to you, the personalization and convenience is worth it.
[00:20:26] Paul Roetzer: Great. You will absolutely benefit. , but if you’re someone who’s more private and you don’t want to have the HA having access to that data and you don’t necessarily trust the providers, then you’re not going to necessarily see those benefits, but you’re making that choice in the other cases. Where you’re not allowing it, it is a business because you’re in highly regular industry.
[00:20:45] Paul Roetzer: So you can’t, it’s not even your choice. Like you just happen to work in an industry for a company that isn’t going to allow access to these tools. And when we’re talking about generative AI and predictive applications, it needs access to data to do what it does, um, and to really fulfill the promise and give you the full benefit.
[00:21:04] Paul Roetzer: So I think there’s gonna be this fine mix between, you’re gonna make choices that determine how much you benefit from ai, and then in other cases, your position in life, whether it’s you can’t afford the tools or you’re, you know, you work in these different organizations, um, or maybe you’re in a country that doesn’t have access to these, is shutting off access.
[00:21:22] Paul Roetzer: So it’s gonna be this mashup between, you know, making choices and then having choices made for you. That’s gonna dictate how much you benefit from ai. So we, what you’re gonna see this play out as, People online will, will start questioning, oh, AI’s really not doing what it was supposed to do. Mm-hmm. , we’re not seeing the benefits.
[00:21:39] Paul Roetzer: I don’t really see it like everybody’s, you know, hyping this up and it’s like, well, you just may not have the same distribution of the benefits for these reasons we’re talking about. And so that was, you know, again, really my goal here was to help people realize you may have a totally different perspective on AI because the benefits and features may not be as distributed to you as they are to other.
[00:22:00] Paul Roetzer: And the people who have all this access and fully understand the technology and embrace the fact that they’re given some stuff up in the process, they may reap more benefits and rewards from AI than you do. So it sounds
[00:22:14] Mike Kaput: like one step is kind of figuring out your level of comfort with kind of the data you’re willing to give up in exchange for those benefits that AI provides.
[00:22:23] Mike Kaput: What are a couple other steps we could leave the audience with before we shift topic?
[00:22:28] Paul Roetzer: Yeah, so I, I outlined three things at the end of that post, and again, we’ll link to the post in the show notes, but the first was to be curious and explore it. So seek the knowledge, whether it’s books, podcasts, documentaries, courses, TikTok videos, or just testing like chat, g p t, whatever it is, you have to understand it.
[00:22:46] Paul Roetzer: In order to maximize the potential positive impact on your personal professional life. So like curiosity, go find more knowledge about it. Listening to podcasts, like if you just listen to this every week, that’s fine, as long as you’re doing something to be like staying at the cutting edge of what is possible.
[00:23:00] Paul Roetzer: The second is the data. Uh, your family and your company are willing to give up in exchange for convenience, personalization, and all the other benefits. And the last was to consider your company’s policies and principles around ai. And do what you can to ensure a balance between innovation and the responsible application of those technologies.
[00:23:16] Paul Roetzer: So it was kind of a, a, it was a hard post to write, honestly. And by the way, I used zero AI in writing the post because I was trying to blend between personal and professional lives and trying to find like principles that applied to both. Um, so yeah, like read the post. I’d love to hear from people, you know, once they’ve had a chance to take a look at it or listen to this and, you know, get your thoughts on it cuz so.
[00:23:39] Paul Roetzer: We’re just putting ideas out into the world just to start conversations. And this was definitely one of ’em where I was like, I don’t know how people are gonna react to this idea. Um, but I’m, I’m always happy to hear feedback from people and, and, you know, take a different perspective on some of this stuff.
[00:23:54] Paul Roetzer: Well, our third main
[00:23:55] Mike Kaput: topic today really does show why. You putting these ideas out there is valuable because, you know, these aren’t just coming out of nowhere. They’re being informed by other trends and other observations that people are putting out there. And one of the big ones that came out very recently was about AI and productivity gains.
[00:24:15] Mike Kaput: And so, you know, when we talk about AI generally, we often hear. That there’s gonna be these wondrous productivity gains produced by AI that will give us back more time in our days and in turn make us feel less busy and more fulfilled. You alluded to this, this utopian outlook on AI’s impact. And there’s no doubt that there will be really valuable productivity gains.
[00:24:38] Mike Kaput: Like for instance, venture Fund ARC Invest predicted that AI could actually boost the productivity of the average knowledge worker by 140%, which would create somewhere around 56 trillion in value globally. However, this isn’t the only narrative or perspective out there, a new article from an organization called The Center for International Governance Innovation Challenges the idea that AI will liberate our time, and it goes so far as to call the AI productivity narrative a lie.
[00:25:11] Mike Kaput: So they point out this fact that. History has shown us that gains inefficiency or productivity as a result of new technologies rarely liberate those already overburdened in society. Instead, new technology, aside from delivering productivity gains, also creates new expectations and norms, which then heightens.
[00:25:32] Mike Kaput: Our standards of the amount of work we should be doing to attain these new standards and norms. So they this, they say, this is known as Parkinson’s lodges. The idea that work expands so as to fill the time available for its completion. You know, like we’ve all experienced how meetings scheduled to last an hour, regardless of what you have to discuss, are probably going to.
[00:25:55] Mike Kaput: To be an hour long. So their whole idea here is that throughout history we’ve seen these technological advances that have no doubt improve productivity. But it didn’t mean that because we’re doing 20 less hours of a certain type of work, that we just have suddenly 20 hours of new free time. In fact, they show throughout history things like the invention of household appliances, you know, they made us more productive.
[00:26:20] Mike Kaput: But instead of reducing, you know, at that time, uh, a long time ago when they were invented, instead of reducing a woman’s labor at home, sociologists actually found that it expanded the amount of hours spent on cleaning, food prep and child rearing because our expectations for the final outcomes we wanted were so much higher.
[00:26:41] Mike Kaput: So, you know, you would think it’d be great. A washing machine, a dryer would allow you to do laundry more frequently. But it actually made us expect laundry should be done more often, therefore creating more work. And so they actually point out that the same thing is happening with ai, that it’s not going to free us up for more meaningful tasks or more leisure.
[00:27:02] Mike Kaput: It’s simply going to change the expectations. And expand the work that we’re expected to do. So Paul, I know this resonated with you and you wrote a LinkedIn post about it and you said, what I’ve learned since is that AI on its own won’t extend time for me or anyone else. What made you come to this conclusion and kind of flag this topic as important?
[00:27:26] Paul Roetzer: I don’t, I, I’m not sure if I’ve ever publicly talked about this before, so that was like part of my motivation was to again, kind of put into words something I’ve thought about for a really long time. Um, but I definitely haven’t written about it, in a sense. So the, the story I told was basically that, like when I started studying artificial intelligence in 2011, there was actually a motivating factor for me, which was to figure out how to extend.
[00:27:52] Paul Roetzer: So I’d become, you know, I had started my agency in 2005 and I worked a lot of hours. Um, you know, so before my daughter was born in 2012, You know, I had, I had like six plus years of, of a startup basically. And you know, I was working 70 hour weeks pretty consistently, seven days a week. Now, I loved what I was doing.
[00:28:14] Paul Roetzer: Like it didn’t bother me per se, that I was doing that. But once I had a child, everything changed. And so I remember like the day my daughter was born, I made a promise to myself that I was done working nights and weekends, that I was gonna find ways. To live a more fulfilling life and to not be consumed by work and, and miss my, my kids.
[00:28:34] Paul Roetzer: Um, so then we had a, our son was born 15 months later, so we have, we have two children and I’ve been very conscious of time ever since. And the thing that always, I always found fascinating, I’ve actually gone like down the rabbit hole trying to understand time from a physics perspective and just from like, what, what is time and why does it always move forward?
[00:28:52] Paul Roetzer: And like all these like questions around time, why does it even. And so it’s like this mashup of fascination with artificial intelligence and time, because as we get older, I always wondered, why does time move so fast? Like when I was a kid and I was in grade school, it felt like those years went forever.
[00:29:09] Paul Roetzer: Like I remember being in fourth and fifth grade and thinking like, am I ever gonna make it to sixth grade? Like time was just so long, and yet it was the same 24 hours in every. And so I, I don’t know when I kind of like arrived at this and it was a documentary I watched, just like self-reflection, but I realized that the older I got time moved faster because I was busier.
[00:29:29] Paul Roetzer: So when you fill your time with like nonstop work and nonstop activities, it just goes really fast. So anyway, I said there’s, I knew I couldn’t get more than 24 hours out of a day, but I thought it might be possible to slow those 24 hours. . So in like 2000 12, 13, 14, when I really started studying artificial intelligence and I realized the potential of it to drive efficiency and productivity, I, I started looking at AI as a path to extend time for me personally and for other people in my life.
[00:30:00] Paul Roetzer: Uh, again, not gonna get more hours out of the day, but maybe I could get more out of those. And so if I didn’t have to, you know, work the 70 hours, but still could achieve the same outcomes, you could still create the same value for people, but do it in less time. And so I’ve, I’ve felt that in some ways I, I wouldn’t say I’ve been naive about it, but part of the pursuit of AI for me has been to try and do this, to achieve this.
[00:30:24] Paul Roetzer: And so then you read this article and it’s like, yeah, it’s a lie. Like it’s false. It’s not gonna, And so I thought the article was really well written, and I thought that the argument being made was very clear and I liked the historical perspective, but to me, the thing that I arrived at, at the end is, is yes, like it is a choice.
[00:30:43] Paul Roetzer: Like this kind of goes with the law of, of uneven AI distribution. You get to choose whether these benefits come to you or not. Hmm. So if you’re a, a business, and this is my fear, is especially with public and traded companies and startups, they’re just going to give you more work. Yes, we’re going to, we’re gonna get 20 to 50% efficiency gains, and you can do 20 to 50% more now.
[00:31:05] Paul Roetzer: So you’re still gonna work 60 hours. You’re still gonna, and so if we don’t, as leaders, focus on helping people live happier, more fulfilling lives as the result of ai, then it’s just another technological revolution to make people work more. And so I think a big thing for me to make sure that this plays out the way I, I, I want it to for all of.
[00:31:29] Paul Roetzer: Is we have to have these conversations that AI isn’t just another tool to get more out of every hour of every workday. It’s we can achieve more produ productive professional lives. We can create more value in our professional lives, but in an ideal world, we get to redistribute some of that time back to ourselves for our own health and wellness and just enjoyment because the time is.
[00:31:55] Paul Roetzer: limited And again, like, I’m not gonna get into a bunch of personal stuff, but like when I was in my late twenties, I, I lost two people. I lost from my best friends and I, I lost my father-in-law at age 52. And so my friend was 29, my father was 52. And so like, that was a few years before my daughter was born. But my awareness of time and how finite it was, was very acute at a very young age.
[00:32:19] Paul Roetzer: Right after I started the. . And so again, for me, this isn’t like a nice to have, it isn’t some feel good thing like, oh yeah, let’s get some in. This is like real, like this is why I’m doing this. Like I want to enjoy my life. Like I don’t want to spend it working 70 hours a week and I don’t want the people that work with me to have to work 70 hours a week.
[00:32:36] Paul Roetzer: So that’s why I said at the end like this is, we have like one chance, like AI is a generational opportunity to give ourselves the greatest gift of all, which is more time. The one thing we can’t get other. But we have to make a choice to do that. It is not going to do it on its own as the article highlights.
[00:32:54] Paul Roetzer: It will just consume more time. We’ll do more, but we’re gonna keep working the same hours. And so I think in the future, you’re gonna, you’re actually gonna make choices about where you work and the kind of company you build based on your acceptance of like, okay, this is enough profit, this is enough time into things.
[00:33:12] Paul Roetzer: Let’s really have a balance in our lives. Let’s really actually be able to, you know, enjoy the other things outside of work, um, because AI enables us to do that, but it, it’s gonna be a hard choice, especially in the current
[00:33:24] Mike Kaput: economy. That’s really fascinating. So it sounds like my read on the article is, you know, we’re all in agreement with the article diagnosing the problem, it sounds like, and the art, but the article itself, you’re kind of going beyond that and saying, you know, cause the article just says no, this won’t happen.
[00:33:41] Mike Kaput: Yes, yes. ,
[00:33:42] Paul Roetzer: it doesn’t present it as a choice. .
[00:33:44] Mike Kaput: Right. And so that’s a key and important distinction here though, is you are, That we do have a once in a generational choice, at least with how we use artificial intelligence and develop it or the benefit of humanity versus, like you said, just another
[00:34:00] Paul Roetzer: productivity tool.
[00:34:01] Paul Roetzer: If, if, if we don’t have a choice, then my work in AI is for nothing. Like I’m not doing this just to like help people save a few hours on marketing tools or like find the next cool marketing tool. Generates some more leads. Like that is not why I’m doing what I’m doing in ai. So if, if there is no way to actually benefit humanity or us individually , um, as a result of ai, then, then there’s no purpose in me doing it.
[00:34:30] Paul Roetzer: So I, I have to believe that. Right. Because otherwise my mission with AI is useless.
[00:34:36] Mike Kaput: Yeah. And I mean, I. Obviously agree with you on that. I do think it is worth believing in an evidence points to the fact we do have choices on how we can develop this technology. We’re having those conversations actively each and every day and week in the industry, uh, about the choices that could be made or should be made in this space.
[00:34:59] Mike Kaput: All right, so let’s jump into a couple rapid fire topics as we wrap up the podcast here. So first up is, what has been dominating the news in any tech or startup circle or finance circle that you may follow online? So Silicon Valley Bank, you know, over the weekend, uh, everyone was kind of watching, holding their breath, thinking there was a possibility that SoCon Valley Bank, due to a couple factors, could actually.
[00:35:27] Mike Kaput: Cause a larger industry-wide bank run and or contagion, the government stepped in to back the depositors, which seems like a positive mood to quell some of the fear around what is going on at that bank. Um, you know, we’re not gonna get in some deep technical analysis of what’s going wrong and what’s coming next.
[00:35:47] Mike Kaput: But we do have friends there. We have partners who bank there and we’re monitoring very closely. This story and anything related to it, given Silicon Valley Bank’s importance in the tech world and startup world, and you know, as part of. What happened with the bank? You know, kind of a run happening on Silicon Valley Bank.
[00:36:07] Mike Kaput: It has a potential downstream impact on innovation in the US because we were suddenly talking almost overnight about the fact if this bank goes down and depositors are not taken care of or backed, we could set back US innovation in the startup. Back a decade. So Paul, I just wanted to get your kind of top of mind thoughts.
[00:36:27] Mike Kaput: Obviously situation is still developing as of this morning, so keep that in mind. But what did you think about this as a whole when it comes to AI tech and innovation?
[00:36:38] Paul Roetzer: Yeah, it was, I mean, it was just really a wild four or five days. Uh, and again, it is still kind of ongoing. Um, but for people who don’t know what we’re talking about, I guess it’s good to just real quick clarity.
[00:36:50] Paul Roetzer: So basic premise, the F D I C ensures $250,000 of deposits in a bank. If you’re a venture backed company, let’s say three weeks ago you raised a 10 million series a round and you put that 10 million into Silicon Valley. As of like Sunday morning, there was a chance you were getting 250,000 of that 10 million that, that’s the basic premise of what happened.
[00:37:14] Paul Roetzer: So they had like 160 billion in deposits and only about 3% of that was covered by F D I C. So all these venture capital firms and VC funded startups, Had all of this money, billions of dollars, that was like at risk of being wiped out, or at least at risk of only getting, you know, 60 cents on the dollar, 50 cents on the, so basically people did nothing like these.
[00:37:38] Paul Roetzer: Depositors just assumed their money was safe and they could just lose everything. So we were at risk of wiping out all of these startups Now, People would be like, oh, it’s just a bunch of rich people. No, no, no. They, that means they couldn’t pay their employees . It means they couldn’t pay their vendors and their partners.
[00:37:53] Paul Roetzer: Like this was a, this is a trickle down effect in the economy. So it was very clear to everyone that the, the, the government had to find a way to make these depositors whole, not to bail out the executives who made bad decisions, or, you know, the investors like, they’re gonna lose everything. That’s fine.
[00:38:09] Paul Roetzer: Except for the executives who took some bonuses, ironically, like a few days before this all. Um, but anyway, uh, so the, this was a very important thing that they stepped in and did this. Now there’s a, there’s a lot of other steps that have to happen, but yes, like you, you could, if startups and venture capitalists don’t think the, that US banking system is safe to put money in, then what are they supposed to do?
[00:38:33] Paul Roetzer: So the government had no choice but to step in here. It was the right decision, um, to be continued. But I think that for us, you know, it was. I was watching it on a number of levels, but certainly one was the impact it could have on innovation and the advancement of AI in the United States in the near term.
[00:38:54] Paul Roetzer: And you know, I don’t think we’re out of the woods yet, per se. I think there’s some other. Um, downstream effects that are gonna take place, but overall, what, what had to happen happened over the weekend. Yep. And hopefully this is not a story we have to keep going into, uh, in future weeks. . Yes.
[00:39:12] Mike Kaput: We are hoping that is the case.
[00:39:14] Mike Kaput: And you know, obviously our, our thoughts are going out to all of our audience and people that may be affected by this say, is not a fun
[00:39:21] Paul Roetzer: time . Yeah, I’m sure. I can’t even imagine what people are going through over the weekend, like just wild. So, yeah, definitely. Um, glad that everybody’s kind of in a better place Monday morning than we were on Sunday morning, and hopefully everything, you know, works out for everybody.
[00:39:39] Paul Roetzer: All
[00:39:39] Mike Kaput: so our last rapid fire topic before we end the podcast here comes from a tweet from a guy named Tristan Harris, who is pretty well known as a former ethicist at Google, and also runs a, a premier or one of the premier organizations for, um, kind of ethics and technology called the Center for Humane Technology.
[00:40:00] Mike Kaput: And he actually tweeted about what we might call the danger. Of the race to put chat g p t into everything. So, um, Tristan tweeted about saying quote, the AI race is totally out of control. And then he goes on to reference, Snap, the parent company of Snapchat, their new my AI chat bot, which is baked into Snapchat and that obviously any user on the platform, especially given it’s a very young user base, can chat with, converse with, and talk to the same as they would a human.
[00:40:35] Mike Kaput: And so he goes on to say, Here’s what snap’s AI told one of their, um, employees, his co-founder, who as an experiment signed up to the service at pretending that they were 13 years old. And they talked about some really sensitive topics where the chatbot told her things like how to lie to her parents about a trip with a 31 year old man and how to do some other things that.
[00:41:04] Mike Kaput: Venture to say many parents would not be comfortable with a 13 year old learning from an artificial intelligence tool. And really, Tristan goes on to flag this as saying, Our kids are not a test lab for this stuff. We’ve talked about how these tools work. Not only can they be extremely convincing, they don’t always have guardrails that say a human or a parent would have, and every interaction they’re having with the user base they are learning from.
[00:41:33] Mike Kaput: So, Paul, what did you make of this? I know you flagged this as pretty something pretty important, uh, in your eyes.
[00:41:39] Paul Roetzer: Yeah, I, I mean, it’s terrifying. So I actually did a class last year on the psychology of social media for kids, for my, my kids’ school with a psychologist and the, the school president. And we talked about stuff like this, like the, you know, not, not at this level, but like there was one of his tweets that every Snapchat user now is an AI friend.
[00:42:00] Paul Roetzer: Unlike human friends will always respond as always available and is always friendly and understanding. Many kins will end up leaning on their chat bite for emotional support more than their human friends on a chatbot that we don’t know what it’s gonna do or say. Right? It has emergent capabilities that aren’t predicted in the labs haven’t been tested out.
[00:42:17] Paul Roetzer: So it is, it is so beyond irresponsible of brands to be infusing this in with no knowledge of how it’s actually gonna work. Parents don’t understand the technology, the kids don’t understand the technology, and apparently SNAP doesn’t care about the TE technology. So, I, it, it’s just, yeah. I mean, this is a horrible example from a human perspective, but it is representative of things that, like even B2B software companies have to be aware of.
[00:42:44] Paul Roetzer: Like right, if you’re infusing technology into your platform that you don’t fully understand what it’s gonna do and if it could go haywire, um, that’s on you. Like you have that responsibility to be putting these, these tools into. Platforms that you actually understand how they’re going to work and what the potential negative implications are.
[00:43:03] Paul Roetzer: And I don’t think we’re there as a, as an industry right now. So, yeah, again, like , there’s all these shiny objects in AI and they’re, they seem really cool and it’s like, oh, let’s just build this in cause there’s an API to do it. And there isn’t enough thought going into the responsible application of this stuff.
[00:43:22] Paul Roetzer: And I just, I feel like we need to really quickly move forward as a business, uh, world and as a society and start having harder conversations about this stuff.
[00:43:31] Mike Kaput: Yeah. And one other important point that we have noted on previous podcasts, but if you are new to kind of our show and our angles on this step, there’s.
[00:43:41] Mike Kaput: Government body breathing down the neck of any tech company today that could change saying you can’t do this. I mean, I don’t even know if many regulators or lawmakers even understand at a basic level what these types of models are
[00:43:57] Paul Roetzer: capable of. No, and that’s, I saw a couple of tweets over the weekends, like where’s, what’s the F D I C of the AI world right now?
[00:44:04] Paul Roetzer: Yeah. So when stuff goes haywire, like happened at Silicon Valley, Who’s stepping in to fix stuff? Nobody because they don’t understand it. And we’re, we’re years away from anything like that. Any meaningful governing body that can actually protect citizens. Um, which is what worries me as quick as this stuff’s moving and as rapidly as people just like jump on it, start building it into products and whatever, and they have, they’re not even thinking.
[00:44:32] Paul Roetzer: This stuff more or less taking action on it. So yeah, I mean that’s why I think what we’re trying to do here is, um, matters and, you know, we need more people not only aware of the issues, but like understand it and can start advancing the conversations in their domain and in their industries. Because, you know, you and I aren’t gonna move the needle on our own.
[00:44:53] Paul Roetzer: It’s you, you just need to get the information in the hands of smart, motivated people and, and hopefully enough people pick up the ball and run with it so we can make some progress on this stuff cuz it’s very concerning to me.
[00:45:05] Mike Kaput: Yeah, I think that’s a good ending point for today’s episode cuz hopefully that’s the entire point of this podcast is to communicate this information, make people aware of these issues because really, Like we’ve talked about across all of the topics today, it’s really in our hands to do with what we want and to figure out on our own.
[00:45:24] Mike Kaput: So the more you know, the better equipped you are to do that.
[00:45:29] Paul Roetzer: There we go. Covered a lot of ground, some big issues today. Well, we appreciate everyone listening. Um, yeah. We’ll be back next week, same time, same place. I think I’m. Am I traveling next week? I know I’ve, uh, I think might be a couple of trips, a couple of talks coming up, but we may be recording on location next week.
[00:45:46] Paul Roetzer: I dunno, . Alright, well we’ll talk with y’all next week. Thanks for listening in.
[00:45:51] Paul Roetzer:
[00:45:51] Paul Roetzer: Thanks for listening to the Marketing AI Show. If you like what you heard, you can subscribe on your favorite podcast app, and if you’re ready to continue your learning, head over to www.marketingaiinstitute.com. Be sure to subscribe to our weekly newsletter, check out our free monthly webinars, and explore dozens of online courses and professional certifications.
[00:46:13] Paul Roetzer: Until next time, stay curious and explore AI.