73. The AI Impact Episode

Jonathan: Welcome to The Difference Engine, the show for tech founders, investors and innovators. Paul: What are we talking about today? Jonathan: Well, today we're looking at the topic everyone is talking about, and few seem to be able to grasp the impact of ai. Paul: Yeah.

And we're gonna channel Steve Jobs and hopefully forgive the English here. Have you think different as we look at some historical context to ask what can vacuum cleaners teach us? About the use of ai. Jonathan: Also some up to the minute AI marketing advice, which we have allegedly co-created with the Mighty Analysts at Garner.

You may be surprised at how much we agree with them. Paul: For ones, well, first we attempt to chart a course through the SAS apocalypse, which many see as AI's first legacy Jonathan: we've mentioned before. In the US the venture capital has seen its allocation shift somewhat. If you look at premier firms like Sequoia Capital and, uh, Andreessen Horowitz, for instance, their portfolios are increasingly dominated by AI investments.

It feels like the classic SAS playbook. Is losing momentum. Paul: Yeah. SaaS has obviously defined the last two decades in, uh, tech companies serving different types of applications like Salesforce, uh, early on.

And then, uh, slack, uh, uh, and, uh, part of Salesforce now and Shopify have built entire categories around cloud delivered. Software, but those markets are now maturing and most new SaaS startups are only incremental improvements, uh, within the existing categories rather than category defining breakthroughs. Jonathan: Yeah, exactly. The, the 2020, 2010s, if you can remember that far back, were, were dominated by SaaS category creation.

I mean, if you could define the category, you could dominate it, identify a problem, build a strategy, a narrative. Before building the actual product. Paul: Yeah. The problem with AI is, is very different, uh, and.

To date all the AI winners that we've seen, the early AI winners, if you will appear to be inherently horizontal, when companies like Open AI or anthropic release systems such as Claude or Chat GPT, um, or even new model capabilities like Claude work, the impact is immediate, but it's also across multiple software categories, customer support. Coding tools, marketing platforms, knowledge management, you name it. To me that's the issue that's causing the slide in these highly specialized vertical SaaS valuations.

Jonathan: Why should we care? I mean, clearly a whole lot of value dis disappearing out the market is, is an issue. Um, but from our perspective, all this makes category ownership much harder. If your startup is built on top of OpenAI, Andro or Google DeepMind, the underlying intelligence layer is controlled by somebody else.

Now, that of course, limits how defensible your category really is. It's, you know, looking back, it's like the, the early days in PC software, um, you know, a whole load of application. Uh, providers appeared, um, but they were building to PC dos or, or later to the Windows operating system, uh, which was ultimately controlled by guess who, you know, Microsoft, the, the Redmond Giants. Now Redmonds was quite happy that third parties built applications that encouraged more people to buy copies of the operating system Cash Cow.

That that wasn't until the revenue stream slowed on the operating system. And then Microsoft, of course. Did what anybody would do. It slowly but surely bundled all the third party app supplies right out of existence.

Paul: Yeah. And of course, um, this has been noticed by our friends in the VC and private equity world, uh, and have they started adjusting how they evaluate startups, which of course, as a reminder, mostly. Are, um, valued on future earnings and not what they're doing today. So if those future earnings start to look even a little bit foreshadowed, if they have an issue, um, during the SaaS boom, which, which they enjoyed for, for probably a decade or so, the sort of questions you ask yourself was, can this company dominate?

X SaaS category. Uh, now the question's more like what proprietary advantage that they also say competitive moat. Does this company have, is it data, is it distribution? Is it their models?

Or as, um, we found out from the dinner we hosted this week. Is it their brand? And I'd also add that there may have been room back in the day for five or six players in this exactly the same category, uh, which made fast following a very valid strategy. You know, you didn't have to be the absolute dominant.

There's plenty of money, uh, fighting for the crumbs from the table. But now unless there's genuine differentiation, life has become much, much harder. Jonathan: And as you, as you mentioned, moats, um, the reality is, is the nature of moats has changed too. Uh.

You know, if you think about it, SaaS companies relied on workflow lock-in, you know, combined with a amazing sales teams. You know, we've had some of that experience on this, uh, on this pod in the past. Uh, and of course those amazing networking events, which, uh, we all remember. Um, with great fondness, uh, by the way, if we could remember them.

Um, and there was also the, the, the, the, the cost of switching, which was often very high. But of course, what people did as they were building those companies was ensure that there was integration into ready-made ecosystems. Now, AR companies, AI companies are different. They rely on different things.

They rely on training data, um, the availability of compute and, and memory access, um, and model context and or quality. Paul: You know, we are generalizing a little bit here for this, talking about AI companies, but one thing we've noticed is the speed of category form formation is a step change. So a SAS category could take years, potentially decades, uh, a decade or so to emerge. And from an investor point of view, this is lovely.

Uh, this meant many years of compounding growth and optionality. On when they could exit, they could ride something that was still growing. And year two, year three, you secondary share sales, uh, which the public never got a sniff of to return their funds to their shareholders, et cetera. They had time on their sides, but AI categories appear and saturate in.

You know, months or even weeks. Uh, let's look at AI coding assistant GitHub copilot. Um, now ironically owned also by Microsoft, uh, which was a snip at over $7.5 billion that practically redefined the category overnight.

Um, but there's new players that, you know, you've now, we've got lovable, we've got Rep Rept, we've got code, uh, codex, we've got, um, open Claw and the whole Vibe coding revolution has sped everything up. Jonathan: You know, we've noticed that. This sort of disruption leads ultimately to, to what we term category inflation. Um, if you think about it, if, if you're living in it, uh, every day at the moment, uh, every new startup now describes itself as a, an AI platform.

And, and we know the reality is, is many of them are just thin wrappers on top of foundation models. The sort of thing we've seen before is there's been paradigmatic shift in enabling technologies in it. You know, when I, um. Weighed around in, in the world of tech m and a, uh, during most days.

Um, I've noticed that this leads to a curious m and a challenge, uh, particularly in Europe where individuals believe they can outta the blue, create technologies that transform the SaaS experience. Um, they believe that they're, they're consulting. Value add channel will result in others play paying huge sums for their apparently game changing ip, uh, which of course they've created in just a matter of weeks. Sorry, Europeans.

No. You know, stuff like that might occur as an outlier in Silicon Valley, whether layers of relationships and previous connections behind that decision to buy rather than to build anyway. Um, lots of this happening, but I digress. Let's go back to that wrapper problem.

Paul: Exactly. So we're, we're focused on European category creation, and we recently spoke to a, uh, company in the code, let's just say the coding space mentioned by coding, uh, and what it would do to their market. Didn't want to know because they had nice, um, local, regional, current, uh, live relationships with, um, you know, let's just call 'em national. Uh, flag carriers.

So they're missing the point that, you know, the European wrapper problem is real. If a platform provider, let's say Microsoft or OpenAI, decides to ship the same feature as you were relying on overnight, you leisure dis dis differentiation overnight. Uh, and in the US you can see this happening. You know, rapid speed.

You can see Acquihires choke point acquisitions like OpenAI buying open claw, uh, meta buying malt book. Uh, so open claw for a rumored billion dollars, you know, speculated numbers. And, um, malt book for, again, speculative, alleged hundreds of. Uh, billions, who knows?

This stuff happens lightning fast in the US to date. This is not the game that we Europeans can play. Jonathan: Yeah. And, and they, these characteristics, um, just reinforce the, the fact that that venture funding is polarizing between the comparatively rfi, ai space and everything else.

Um, yeah, of course. Including good old SaaS application layers. Um. You know, high capital flows into and between foundational AI companies and infrastructure players like Nvidia, open Air and Anthropic and, and so on.

Um, while the funding for applications is comparatively sparse or, or non-existent to date, Paul: yeah. Seeing a lot of that in, uh, offered companies on the private market, it's ironic because during the SaaS dominant era, it was the application layer. That captured most of the value, uh, and infrastructure, uh, pro providers, you know, they, they, um, they just had to, um, feel like they were commodities. It's now the other way round.

You've got a name for this, I think. Jonathan: Oh, of course. Yeah, yeah. So we, we call this the great software witchu.

So the, the question becomes if SaaS categories are saturated and AI is controlled by a few foundational players, where do the next. Categories emerge. Paul: Yeah, I, I, and again, I think some of that came out from, from the dinner that we hosted this week, but my go, my guess is that, um, AI native workflows will be the new application layer instead of embedding AI into existing SaaS tools. That's, you know, that AI washings, um, sort of discredited disruptive new companies and there could be, a lot of them actually will build completely design and build brand new workflows around AI from the start.

Jonathan: Yeah. And this is, this is. Paradigmatic shift. And that means there has to be a shift in thinking.

And in this case, you know, we think that the category shifts from something based on the idea of software tools, which we've been used to for decades to something closer to digital coworkers. Paul: Yeah, exactly. Um, and not the sort of coworkers we see sort of lamely offered today. So instead of tools that help people in theory, work faster, we're gonna see systems that perform entire parts of.

Uh, their workday autonomously. So coding's the obvious one. Then research analysis, even operational decision making, and it's the operational decision making where the humans are involved. That is the crunch point.

Jonathan: The new challenge in, in category design is that you'll no longer just selling software or even. Selling its benefits, you are asking companies to trust an AI with decisions about your business process. Paul: Yeah. And that introduces, dare I say, governance.

Uh, mature themes like oversight and of course trust. And the real category battle here in EMEA might end up being who defines that operational? Layer that sort of governs or manages the AI systems because that's a place where a lot of value can be added or subtracted and real category power can be exercised. Um, one example, which we.

I helped launch in this space is a British company called Oma. It's a agentic commerce protocol. A MP, which stands for Agentic Merchant Protocol, is in fact is, is a, is a category definition, and it creates a new system of record. It's not a platform in the sense of a, of an LLM, of an infrastructure platform.

It's a layer. With a real value to its, uh, ideal customer profiles, who by the way, include customers like Mars Perfect Ted, and several of the well-known brands. It sits above Amazon and the other online shops and gives control back to the brands and the merchants. Who, let's face it, spend billions to create customer pull only in some cases to hand over all of their creation, all of their creativity, all of their brand power to these massive hypermarkets, uh, who then can promptly go ahead and advertise, uh, rival products against them.

Um, and that's why we've got the Oma, CEO Max Sinclair on a future. Sharing some of the lessons about what this new type of category design involves. Jonathan: So I think if we're to draw some lessons from this is, is that the next category might not be dominated by the smartest or as has often happened. In the past, not the smartest but best distributed AI model.

Paul: Yeah, right. It might be the company that best defines how I, how AI actually operates inside large organizations. Uh, and that could be the savior of some of the massive consulting. Organizations who are, um, and I'm thinking of pwc at the moment, who currently are beginning to, and, and maybe aggressively reevaluate their top to bottom structure for service delivery.

Jonathan: Yeah. Well that, that might be good news for pwc, but it's also gonna be, or could be good news for likes of Accenture IBM consulting, Deloitte. Capgemini, cognizant, tartar, emphasis, Wipro, and of course their employees, but only if they can re-skill fast enough. Paul: All right, so today's lesson from Tech History is two paradoxes and a law.

Jonathan: Uh, well, everybody's been talking about agen this and agen that. A lot about what AI agents are going to do to knowledge work. Paul: We're gonna try and look back at work related lessons from yesterday, uh, and see if we can tackle some of the principles of work that, uh, seem spot on. For today, Jonathan: it all starts with a, a classic work of women's history.

A book from Ruth Schwartz Cowan. Definitely worth reading. Um, if you can get past its title, not terribly snappy, um, it's, its title is called More Work for Mother, the Ironies of Household Technology from the Open Heart. To the microwave.

It did win the 1984 Dexter Prize from the Society for the History of Technology, a very August body. Uh, and you know how much I would love mu on the, the history of tech and what it can teach us. Now in this case, what I, what I particularly love. It's that it describes one of the most counterintuitive results in technology adoption.

Labor saving household appliances don't actually reduce labor because we've spent most of our lives listening to a whole load of promises, you know, and concerns about machines or machine adoption, uh, eventually leading to the sunlit uplands of endless leisure for us humans. Paul: Uh, it cannot be. A coincidence that this paradox about labor saving devices not saving labor, uh, is about domestic tasks. Um, and the reason I say that is even the most highly inventive, uh, you know, nations on earth, the Japanese and the Koreans who have got aging populations and real concerns.

Uh, can't find solutions for these sorts of problems. Jonathan: Exactly. And, and, and this, this, this phenomenon, which we described with the very snappy title is now thankfully known as the Cowan Paradox. Now, to me, that cemented Ruth's place in tech history.

Um, I reckon though. That it's pretty useful as a guide to predicting what AI agents will actually do to work. Paul: Are you sure? Uh, I know you love this stuff, but just gimme the TLDR too long.

Didn't read version, please. Jonathan: All right. Okay. Okay.

So take the vacuum cleaner or, or Hoover if you want the Boomer generic or perhaps even now the. Rah for the UK Dyson. Now, before electric vacuums existed, cleaning rugs, uh, if you were flush enough to afford anything resembling a carpet was a huge ordeal. Uh, involved moving furniture, rolling up heavy lumps of textile, taking 'em outside, hanging them on a line, and literally.

Beating the dust outta them with paddles. Right. Paul: That sounds good for your mental health actually. Jonathan: Uh, probably good for your mental health, very good for your forearms and, uh, cheaper than a gym membership.

But if you think about it, it took multiple people to do this and sort of because of that and other, other factors, it happened maybe once or twice a year. Now, of course, that's where the idea of spring cleaning eventually came from. Paul: Right? So just cleaning therefore was.

Infrequent because we're humans and it was inconvenient, uh, it was exhausting. It was time consuming. It was effectively expensive. In terms of labor.

Jonathan: Oh, getting all economic there. Paul: Yeah. Yeah, yeah, yeah. In economic terms, labor was comparatively cheap back then, though, still quite knackering though.

Jonathan: Then electric vacuums appeared in the early 20th century that this new category of technology, um, as. What happens when new technologies do burst on the scene actually required the perfection of a number of innovations, um, you know, from domestic electricity supply, smaller motors, new materials, and so on. But the point was suddenly you could clean the rug or increasingly the fixed. Carpet actually installed because of the arrival of the Hoover.

Yeah. Right. Where it was sitting, no lifting, no helpers. Paul: And that's handy.

Um, because given the time post the Second World War, women were entering the workforce, uh, as, and I mean the non-domestic workforce, what we would call work. So logically, the, the time needed for housework had to be reduced. It reduced. And um, here was the technology ready to deliver this very time sa saving, um, time that could be used.

Like even more value added ta to, or even leisure Jonathan: I had. But that's where you're wrong. It didn't, it absolutely did not. That's actually mostly determined by three things that changed at the same time.

What happened because you got these handy Hoover Dyson vacuum cleaner things was the, the frequency of use exploded. Right? So once cleaning became easy. Society changed, social expectations shifted, right?

So manufacturers of course drove demand for their product based on, you know, the stuff you see coming outta p and g and lever these days, you know, scare tactics about germs, disease, and dirt. So instead of cleaning once or twice a year, and everybody did that, so that was fine. Households were guilted into vacuuming weekly or even. Daily, right?

So you're doing it more. So second, the actual labor got reassigned, right? So think about it. Beating rugs actually required physical strength.

So, you know, men or hard help often did it. Um, vacuuming of course was, I mean, seemed to be and actually was easier. So it got reclassified as lighthouse work, which meant it became the wife's responsibility. Um, how convenient for the patriarchy.

Anyway, third thing that happened was that helpers disappeared. Um, and this was, you know, a product of Post First World War where there weren't that many people around because they'd been killed somewhere in the mud in Europe. Um, and middle class households used to employ domestic servants, couldn't find them or found there was an alternative. So for that level of heavy cleaning, we now have an electric servant.

Um, replace the human. And, and of course, if you think about it, the same thing happened with domestic laundry. You know, you, you, you, you used to ship it out for somebody else to do, um, or, you know, the woman of the house would spend hours and hours on the tub working away, cleaning, cleaning the clothes. But then what happened was Launderettes opened so all the capital, uh, was put into putting common law machines in, and that became a sort of halfway.

Way house before everything got cheap enough and small enough, and we started to own our own laundry machines, uh, in the uk. Certainly by the eighties. Paul: There you go. The labor's disappeared.

Jonathan: It actually merely transferred the labor, but not the time taken because the launders mostly house wise, needed to mine their washing and transfer it to dryers. Paul: The Cowen paradox, there was relatively little time saving. The all that happened was the format of the work just moved. And, and in fact, some of the tasks multiplied, I can see the argument you're making.

In fact, um, I witnessed that, uh, just this week. Um, somebody was talking about SEO roles, uh, uh, in a consumer packaged goods companies. And, um, yeah, uh, that work needs to be done. It's just less important and, um, probably broken up into different tasks.

So looking at keywords, you know, endlessly updating HTML pages. This is not a job for humans. Um, and thankfully with ai, the nature of this work is actually changing this, this sort of shift changes the very meaning of. Quote unquote work and reduces the need for certain job roles in paid employment just like it did in households.

Jonathan: Yeah. That, that also a exemplifies how quickly labor shifts, because there was a point when endlessly updating HCL pages was a job for humans. Uh, you know, just like doing the laundry, a job you really wouldn't want to do. The Camp Paradox basically says the time spent doing the work.

However it evolves basically does not change So empirically time used studies from the thirties through to the fifties show that housewives men were frankly really not doing their share of work then, um, the housewives were actually still spending about 51 hours a week on housework, basically unchanged before any appliances. Arrived. Paul: Right. And, and that's the heart of the, the Cow Paradox.

Labor saving technology doesn't necessarily save labor. It resets expectations. You think that means we can expect to see more pro more production of HTML pages and more human checking of the output, maybe. Jonathan: And that could be a result.

I mean, ca Cow and argued that, that the real effect of technologies is raising the standard of the output. Right? And obviously that's a very subjective, uh, idea. But it, it certainly is true if you think about it, when, when something becomes cheaper or easier, so society demands more of it, right?

So is that beginning to sound like the current AI debate Paul: now for our second law? And this is one that many, many people are talking about in the age of ai, and it's the classic efficiency paradox described by William Stanley Jevons, who published the Theory of Political Economy, uh, while he held the Chair of Political Economy. Guess where? Our alma mater, Manchester University Hurrah.

Jonathan: Yes. And there's a whole building called the Jevons Building named after him. Anyway, the Jevons Paradox. Right back in the 19th century, Jevons noticed that there's a new fangled.

Coal fired engines became more efficient. Total coal consumption actually increased. Yeah, they are efficiency, lowered the cost per unit, which encouraged more use of those newfangled engines. Paul: Same pattern again, induced demand, added roads, uh, and you just increase the traffic.

Uh, improve fuel efficiency. And guess what? People drive more miles or choose bigger Chelsea tractors, as we call 'em in the uk, uh, SUVs as, uh, our American friends call them, and they seem to have negated the theoretical overall reduction in fuel consumption. Uh, due to more efficient engines.

'cause they're massive lumps, Jonathan: they're great big bricks, which you're trying to throw through the air. So what this says is sort of in quasi economic terms, efficiency gains get absorbed by demand expansion and don't deliver reduced effort. Paul: So your argument is that these a a, a AI agents. People are touting are basically the vacuum cleaner of knowledge work.

Well, they certainly hoover up dull tiles like SEO, but should we not expect a life of leisure just yet? Jonathan: Yeah, exactly. They'll make certain tasks dramatically easier, but it won't reduce work. It'll increase the volume and expectations around knowledge output.

Now I think we're already seeing e early evidence of that, and guess what the academics are on it. Um, the Harvard Business Review, a very August journal, which I'd recommend you all to read recently published. A study by Aruna ran and jinky Maggie Yi. Massive apologies for the pronunciation, but we're mere Brits.

Anyway, these two academic stars spent eight months embedded in a 200 person tech company studying what actually happened when employees adopted AI tools. Paul: Now this is interesting and let me guess, the result wasn't less work, was it? Um. It sounds like, uh, if it wasn't that the AI in question was Microsoft copilot, and don't get me started on that one.

Jonathan: Uh, not commenting on the tech. Um, give, give it. I spend, uh, most days working with somebody appears to be half man, half copilot. Um, so yes, yes.

No work reduction. Uh, what they saw was, was three clear patterns. You have to think these things are all deeply rooted in, in human psychology. Um, right.

So what happened was that there was task expansion, task expanded. And explain this in a, in a moment. Um, work boundaries started to disappear because of friction reduction in actually getting the stuff done. And there was an explosion in multitasking.

Um, presumably also driven by the idea that people felt they needed to do more of the AI to keep their basic jobs. So if you think about task expansion. What happened was that they observed that people didn't just do their own jobs faster. They started doing other people's jobs.

Product managers, right? Yeah. Right. So this is what happens when you get vibe coding.

Product managers began writing codes. Researchers started handling engineering tasks. You know, God, God work that previously required hiring somebody new, got absorbed by existing. Employees.

Right, Paul: right, right. So in a way, the, the AI agents here are like the old domestic servants. They're being replaced by, if you like housewives, that's you and I, yeah. Doing a lot more work.

Jonathan: Yeah. But it's also that, you know, it's used, extend that analogy. It's that also the housewives justifying their existence, uh, for want of not knowing what else to do. Um.

Right. So the second thing was that the work time boundaries disappeared. I mean, what a work. Yeah.

What is a work time boundary? I've never, I've never had one in my life. So in this more str, apparently structured environment, the work time boundaries disappeared. And, and this was because apparently the AI tools sort of feel conversational, sort of, sort of human, so workers actually started using them constantly.

Now if you think it's irritating enough, walking down the street with everybody staring at their, their phone and you know, bumping into you. What started happening was was that during lunch, in meetings, before leaving the office, it felt like everybody was just going, oh, one more. Quick, prompt, and you know what happens there because we've sort of been there, um, suddenly it's like 9:00 PM and you're still working. I mean, that to me, that to me that sounds just like the old days down the M four corridor, which is for those people not familiar with the UK was the, with the Silicon Valley of its time.

You know, we used to be just, we'd be, we'd be working the, working away on presentations and stuff and then, you know, we'd be wanting to go home, but then somebody would pipe up. Uh, there's one more thing we, we could put in tomorrow's presentation and we'd feel, we'd feel duty bound to do it. Paul: It was easy about then though, because there were fewer speed cameras. But I digress.

Alright, so, um, so what we're saying about this work, uh, expansion and the task merging, et cetera, is the friction disappears. Uh, so that means the stopping. Uh, for work disappears. And I'm seeing that with ai, um, you just create more subtile, especially, uh, as voice note usage increases the, the amount of inputs increase.

And that's part of the, the appeal I guess, of, uh, the current appeal of Open Claw. Because it works 24 7, it works well with voice commands and it often uses a dedicated server so it doesn't get in the way of anything else or a Mac Mini, uh, while all the other work stays on your other devices. It's actually more work via more channels. Um, and guess what?

There is a backlash because open claw is so autonomous, uh, does so much coding that proper techies, uh, are now calling it inefficient. They would prefer to see. Uh, proper old school, API calls, which means they can oversee them, but, uh, you know, they, they, they may not win the argument 'cause it's defenders open. Law defenders say, well, it's cheap.

It doesn't answer back, and we don't need you proper techies. The work has in fact not gone away. In fact, it's increased. But as you said with your, in your previous two examples, the format has changed.

Jonathan: Yeah. And it's transferred and you know, that's classic Karen paradox behavior. But the third pattern that, that our two academics found was the multitasking explosion, which we're already on a trajectory for. Come on.

You know, we all dip in and out of applications and, Paul: sorry, I was just on my phone. Jonathan: Yeah, right. So, and you know, are you s alarm me slicing your time? Because that's sort of what happens.

Um, and, and, and in this case, in, in this world of AI workers started turbocharging it, you know, running multiple AI agents simultaneously and, you know, it's like, guess believable. Guess what? Old projects were revived because AI could handle parts of them. In the background, you know, stuff they thought had come to a dead end on, I'll just give it to ai, it'll work on it.

It comes out with something Paul: that could create the feeling or the illusion, if you will, of momentum and progress. Jonathan: Yeah. Yeah. Right.

But we all know what the reality is. Uh, and they confirmed this. It's a, it's cognitive overload. Um, and, you know, the, the researchers ultimately concluded that AI increases the.

Intensity of work, you know, multiple copies of communications on email, chat, and other channels. Plus more checking because there are so many channels and tasks being attempted. Paul: This sounds like we need a new social contract, maybe. Um, 'cause AI doesn't contract the work, it actually amplifies it.

Um, you know, one could argue the productivity goes up, but so do expectations maybe even. Faster, uh, and clearly once again, the time spent working does not reduce. Jonathan: And as you would expect, because we're all about category, there is a tech category implication here. From a category perspective, I think we may be thinking about AI agents the wrong way.

Paul: Do tell how so? Jonathan: So everybody is assumed. The primary value proposition for AI is labor reduction. And you know, those are all scare stories that are appearing in the media at the moment, but historically that's almost never what so-called efficiency technologies actually deliver.

Even if people and they still are, are choosing to believe otherwise. Paul: Well, that's a bit of a shocker 'cause the real category that everyone's talking about or seems to be, uh, talking about AI automation. Jonathan: Yeah, right. But actually it might be AI driven output expansion.

And that will be tools that let organizations attempt things that were previously impossible because the labor cost was just too high. Paul: That sounds like a productivity win. And, and, and so if this is true, um, that it's all about AI driven output expansion, the winners here would be, won't be companies that reduce headcount because as we've seen empirically, this just doesn't happen. Uh, and even if there appears to be a lot of attempts, um.

To boost share price by promising AI implementation that will, will not go down, right. Jonathan: That link between AI implementation and headcount reduction, you know, does appear to be a fashion, um, designed to boost share price, but. We think that the winners will actually be the ones that enable companies to raise the ceiling of what their teams can produce, you know, to use the title of a great book to think bigger, if you will. So this brings us to our third and final law, Paul: and possibly the easiest one that everybody knows, right?

If they don't know what it's called, they know it. Jonathan: It's actually called Parkinson's Law, where. Work expands to fill the time available, which clearly already applies here. Paul: Yeah, absolutely.

Work does expand to fill the time available. In other words, AI isn't the end of work or even, even its reduction. The work's increasing and, and what is classically seen as as work. It drifts, it scope creeps, it starts to include nice to haves.

It basically increases. In fact, we've, we've both witnessed, um, machine aid, busy work, uh, within ai. Uh, it's all too easy to disappear down that rabbit hole, um, doing something because you can, not because it's necessary. Um.

Now, is that all bad or is that valid experimentation? Jonathan: Well, we will see, but I think it's just the beginning of a much higher bar for our expectations of ai, at least for those that still have or aspire to have white collar jobs. For those of us who have been in, uh, uh, tech for, for a long time, there's been, uh, one constant, and that is the. Appearance of advice from the world's biggest tech research outfit, Gartner.

And, um, we have, uh, noticed some things that have been said by them, but what really caught our eye is, um, double your PR budget now says Gartner. And we think this is all about a massive shift, um, in how people, um, think about knowledge. And this is about going from offer to discover. Paul: This is, uh, a marketing strategy session, so you might need to listen and take a few notes, uh, get ready, especially for the five fast tactics that we're gonna drop at the end.

So, yeah, you're right. It's clear that the nature of B2B marketing is changing. Uh, customers are moving from, uh, accepting offers, uh, and, uh. Into more of a discovering solutions proactively and solutions to very specific issues.

The likes of which they probably, we spent a lot of time researching before. They don't need to do that anymore. And the upshot for category designers and brand owners is the places that you need to show up have changed fundamentally Jonathan: US. Talk you through this in detail and explain why you need to take action now.

Or risk losing your category. And the extraordinary thing about this is for once, for once, possibly the only time, possibly we will be in violent agreement with the high priests of tech, the ancient Drew. Its at Gartner Paul: here. It is the least likely possible headline of the year.

Drum roll. LLMs will drive a two x increase in PR and earned media budgets over the next two years, says Gartner Jonathan: Shock. Horror, but unsurprisingly, given our joint background in category marketing, communications and indeed tech pr, um, we agree with the Yes, we do agree. You heard that Now I'll just say it again.

We do agree. We agree, we definitely agree with the sentiment expressed here, uh, by the, and let's face it, they're very big 11 billion market cap industry analysts at Garner. But, but, but, but given that market cap. Was nearly double what it was not so long ago.

Uh, these are the guys that predicted IBM would beat Microsoft in the operating system was and rather overestimated 3D printing. So we are a little cautious Paul: indeed, we are still, um, surprised about the agreement, how magnanimous of us. Um, yeah. So the reason that we're saying this though, uh.

History aside is, uh, we're not, this is not the same reaction as some very hard done by hardworking folks in pr, uh, who've been frothing at the mouth at this announcement. No, no, no, no. Guys and girls of pr. Just because the AI is allegedly after your job does not mean.

We need to clutch it. Any straws? We've got this people. Jonathan: Okay.

Do you wanna dig into this a little bit? Paul: Yeah, let's do it. Okay. So, so Gartner's, um, output from their CCO, uh, event, uh, had, uh, three predictions and a prediction.

One, uh, again, just read it out again 'cause it's so good. By 2027, mass adoption of public LLMs. Uh, as a replacement for traditional search, that bit's important as a replacement for traditional search will drive a two x increase in PR and earned media budgets. Oh, Jonathan: that Paul: is good.

Jonathan: Good. Oh, that is good. How many times increase in PR budgets, Paul? Paul: Two x Jonathan: Oh oh.

Hit me one more time. Is that Paul: 300%? It says it's a lot. Jonathan: So what were the sort of key findings then that, that that justified this?

Oh, a rightly good headline. Paul: Well, they're talking about the, um, you know, the highest likelihoods of, of, of things that CCOs or chief. Comms offices would invest in. Uh, and they're talking about how AI powered chatbots like chat, GPT, which they said was up 608% as a source of inquiries and perplexity, which, you know, a mere 262%, but we like perplexity.

Um, and they'd experienced exponential year over year, over year traffic in increases. Meanwhile, they, according to Gartner, traditional search engines were down. Um, you know, a few percentage points, both Google and, uh, Bing. So, so, so they're using data, uh, which we love.

Uh, and you know, it's, um, it seems to be a little bit AI driven, wouldn't you say? Jonathan: Well, this is all very, very, very jolly, jolly good. Um, but as this is the AI era, I think our listeners. Demand a little context.

Listen up the advertising. Yeah. Laterally online advertising has ruled the roofs for decades. Right.

We all know that. In fact, in any other world, that reality would've been an issue for some form of monopolies commission, right? Right. So if you think about it, um, if you think about global search engine share across all devices.

Google has hoovered up close to 90% worldwide since at least 2015, right? So, uh, some later data Statista and, um, stat counter, I think, um, they showed roughly 80, 90, 90 3% share across 2015 to 25. But hey, hey, good times are coming. The arrival of AI has has changed the emphasis from vendor offer to customer discovery.

Now, let me say that again. The arrival of AI has changed the emphasis from vendor offer to customer discovery. Now, the customer through AI can specify exactly what they want from any vendor or any product to a very granular level, depending on the prompt. This changes the game.

Paul: I love it. 'cause the cliche is the customer is king, but the reality is. Google's, um, online, uh, auctions for ads. That was the king, uh, and the rest of us were all peasants.

So, um, now, last week, interesting enough, we attended a special launch of a new. Product, uh, around AI visibility. More of that, uh, later, remember the name Oma. Uh, and at that event we watched an executive from, you know, a true B2C category leader.

This is loop these, the, um, the noise canceling earplugs that, uh, all the rage, uh, for concerts and for, um, sleep Jonathan: amongst young people in particular. Paul: Can you tell that to my other half? I'm sure she'll be delighted you said that. So once upon a time, um, you know, this very erudite head of marketing mentioned it was all about impressions.

Um, frankly, that's how he and everybody in the marketing industry was, uh, measured. And we literally mean, uh, impressions from, uh, search ads and then if they go down. It is a simple solution. You just reach for the wallet or speak to the CFO and because they've gone down, because they're not effective, you go and spend more money with Metro or Google, right?

No longer. So, um, Matt said in his case, and bear in mind, his product is a high consideration, uh, durable good, which you buy once or twice, you know, every few years. Uh, he said that their imp when their impressions went down. That was actually a good thing, which is sort of counterintuitive based on the old way of working because, um, if the impressions were going down and con conversions were simultaneously going up, that was great.

That, that, that meant people coming in. Buying products and moving on. Now, slight issue in that he's always gotta find a new set of customers. But isn't that counterintuitive?

Jonathan: Yeah. So why was that then? Paul: Well, simply because customers were showing up at the point of purchase with their credit card in their hands. And this is, I think a, you know, it's a mid-size 50 to a hundred pound purchase.

People were showing up, much more informed. They knew if they needed, why they needed it, what they, you know, they, they were ready to buy, but. Here's the real question. Where was that information coming from?

And you may be surprised. Jonathan: Oh, yes, yes, yes, yes. The issue is that ai, AI answer engines site. Earned media above all other sources.

And it really doesn't matter what the LLM is. There's chat GPT or Google Gemini or anything else. It's all about earned media. And that would be about 40% according to a zoma on chat GPT and 34% on Google Gemini.

Paul: Yes. So like this is a massive turnaround, as you say. Um. Depending chat, you know, chat, GBT 40% earned media, uh, Google, Gemini, 34% earned media.

But also you have to add to that 12% of, you know, UGC, YouTube, uh, which, which open AI because they don't own YouTube, unlike Google, don't count, uh, for sources. You know, the amount and, and this is where Gartner's right, the amount of, uh, citations, uh, that is the, uh, those are the super scripts that, um, inform the answers that the AI engines or AI answer engines, depending where we end up on that category, are citing that's coming from earned media. It's bigger than any other source.

Jonathan: Before we get too excited, perhaps we should go back to the, the, the source of this double your PR budget strategy. Um, back to the one coats at Gartner. Paul: Okay. You want me to be nasty?

I'll do it. Jonathan: Okay. So No, no, no, no, no, no, no. It's good.

I Paul: find it Jonathan: ironic. So the Gartner wants now to be nice to pr, you know, please respect that. Paul: No, no, no. That's rich because for years, um, Gartner's been stealing a living from, uh, those hard bitten.

P two B Tech marketing folks, they were consuming budgets, flying off execs to, uh, garner Symposias in Jonathan: Bara, Vegas, London. Paul: Yeah. Yeah, yeah. You know, so then you've got, you got, you got your execs out.

Jonathan: That's not a euphemism folks. Paul: Yeah. The expenses for all of that, um, you know, hotels and, and, and meals, right. Et cetera, et cetera.

So a lot of marketing budget just flying off to Gartner. Um, they would also, uh, and this sort of hurts us as category folks, they would dictate your positioning. Um. And often, uh, create what I would see as useless gobby gook, um, to, for insiders, um, magic quadrants.

So, um, you know, there they'd also, and the, these very same magic quadrants would take up time as you know, perfectly good customer references, which could, you know, be used as videos on your website. Uh, which are favors you may want to ask to get, you know, press releases these favors, uh, would be distracting for the sales guys who would, you know, with justification say, no, I'll do it for Gartner, but I won't do it for other marketing things. So that's another bone to pick with Gartner.

Uh, also, um, given Gartner's massive, uh, influence, they distract, um. Product marketing talent, they would spend their entire time getting the message right. Not for the market, but for the Gartner analysts. Uh, and sometimes they'd even think, um, doing real work, IE making their propositions useful for customers, was a little bit below them.

Jonathan: What I, what I was used to love was they, they basically started to demand, they wouldn't talk to normal people. They had to talk to special analyst relations teams. You know, and, and again, once again, you, you're diverting budgets to. Frankly, a bunch of one-trick ponies who, who did a similar but far less measurable role to pr, sort of, you know, the emperor's got new clothes.

Everybody felt they should exist, but nobody could really work, work out why. And I, you know, I'm sure this, and, and the things you've cited there, Paul, are all true and, and frankly who doesn't like a little bit of. Hubris. Um, you know, I hear you on this.

Um, but here, Gartner has been genuinely helpful. Paul: I know, right? Jonathan: Steady on. I'm gonna have to put a bucket of water over here.

This, right, this, this prediction is, is all about a massive change in how to attract customers to your category and to constantly encourage them to find out more about your product before taking that. All important purchasing decision. Paul: And if I may, that that is more important in B2B Tech where you've got sophisticated, somewhat complicated products than if you're selling milk for instance. Jonathan: This is a move for an emphasis, not, you know, just on the offer to a focus on discovery.

They offer and that, and that is a fundamental shift. Paul: One would think that, uh, a lot of, um, your exec time should be figuring out how the discovery process works these days. And it means brands and vendors need a very tight control on how and where they. Are discovered, uh, it might even mean a new category for it analysts like Gartner, but, but enough of that.

What do we conclude from this, Jonathan? Jonathan: Right. Well, we conclude that we're very excited and the reason we're very excited is power has shifted from the producer to the consumer, in this case decisively. Um, so.

People who are trying to sell stuff need to think more carefully than ever how their potential buyers will assemble the information about your category. And of course, your offers within the category, right? So you have to think what will the pain points they suffer. Um, mean that they are looking for, Paul: because they're gonna be the prompts, right?

Like you say, it's rather like category design. Jonathan: So rather than just considering which features might appeal most to the audience, think about how your proposition can be seeded into those channels, which are likely to form the discovery journey of an powerful. AI powered customer, Paul: and we said at the top, and we, we wanna deliver here that, um, we'd have some five very tactical pieces of advice for you. So let's rattle through these.

Okay. So, um, number one today, review and possibly defer your ad spending because you don't know if it's affecting sales. And maybe a good way to do this would be to use AI to determine what the effect of your ad spends are. Jonathan: Secondly.

Audit your AI citations to find out where you're showing up. You know, use a tool like OMA or Peak to track this against your hopefully clear or you know, most definitely soon, much more. Tightly defined ICPs, and of course, that is an acronym for Ideal Customer Profiles, the people that you really, really want to sell to. Paul: At three, continue to tighten up both your ICP definitions and importantly the definitions of their pain points.

Bear in mind, the average AI prompt these days has 26 words as opposed to maybe three or four. For the ad searches that they're replacing. Jonathan: Hmm. So at four, you really do need to pay close attention to the shifting sands of AI visibility.

Are your blog readers engaging? More or less? How many earned articles did your PR create? Is Wikipedia losing to Wikipedia this month?

Paul: Finally, and number five. Please remember, this is all brand new. It's exciting and new, and the relative stability of years of just throwing money. At ad search.

At the ad ad search. Emini is all over. This is the time when marketing people and prs. Can make the most amount of difference to category creators and category creation.

Take back control of your pr, as Gar says, not just us. Uh, and should we say learn to love AI visibility, sisters and brothers, please hit us up. If this is of interest, we'd love to help you think it through. Jonathan: Thank you for listening.

If you wanna learn more about category design, head to. Be categorical.com. If you need help designing and dominating your category, then get in touch.

Contact details are in the show notes.