LXA CEO Carlos Doughty and Godfather of Martech Scott Brinker recently sat to discuss a variety of topics, with the rapid growth and potential of AI being a major theme running through the discussion.
In the first part of the conversation, Scott and Carlos discussed the disruptive potential of AI. In part two, they looked at the potential winners and losers of AI. In this third and final part, they discuss the impact AI is likely to have on the future of learning, as well as why marketers need to focus on the ‘AI of the possible’.
Carlos Doughty: I think it’s really interesting to think about what happens to people's learning capabilities with the growth of AI. We've gone from an era of having to read a book, to really ingest that information in physical format to just Googling it. Now, I don't even need to Google it and read it and understand it. I just ChatGPT it.
So how much of your learning, information storage and understanding do you actually retain? It's about adapting to the new and not trying to force it to the old, but it's an interesting kind of take on things.
There have been examples of ChatGBT being able to take the bar exam or become a doctor, and so on. The problem clearly is not ChatGBT, it's the test itself. It’s that this exam is clearly not a good way in which we now test people if that information is always available to us.
And so some of the ‘antiquated’ ways of being and living, such as turning up for an exam and not being able to use the internet for the exam. Why not? For the rest of my life I'll be able to use it, so what are you testing for?
I'd be interested in your thoughts on how we learn and consume. What does this mean for our ways of operating and really understanding the subject matter?
“It could take humanity up to a whole new level. That's the more optimistic take.“
Scott Brinker: It's hard to say. One of the greatest barriers to learning new things is often when you're in a beginner's state of mind, and you're trying to get help. You don't have an expert person by your side to ask questions. It's just hard to get past the barriers. This is why careers have the moats they do, and you have to put in the hours.
Although Google's started to help with that, the truth is it's still pretty shallow. For example, if you want to figure out how to get Python set up to do something on a Mac and like to experiment with it, have you tried Googling for that shit?
You are off to 20 different sites, some of which are ancient, some of which aren't much use. It's better than what it was, and it has raised the bar, the information was available, but it didn't get us to some really qualitative new place. It’s still a lot of work to learn something.
I see people interacting with GPT and it's almost like a personal tutor that answers any questions you have. It's like this learning cycle.
I guess there's a possible counter-argument, and I'm not saying this is how it's gonna play out. One scenario is, we just get dumber because we don't have to do anything as the AI does it all for us, or the way in which we use AI as these personal tutors helps to advance our own capabilities and our knowledge in domains that frankly, would have never been accessible to us before.
It could take humanity up to a whole new level. That's the more optimistic take.
CD: That’s an interesting take. I think most would agree, take away the administration, take away the stuff we don't want to do that's not necessarily an effective use of our time. Certainly, if you get to play out and question back and forth with somebody that's got a brain the size of OpenAI, it’s pretty powerful.
It's whether you get the benefits you mention if you start to get it to do things for you, such as to write your emails, or use it as a prompt to learn and to think differently. There's a distinction.
For content, ChatGPT could be used as a trigger to start to get into a topic, by asking for 10 quick points on a subject. From a learning perspective, I could agree with seven of them, and add six more of my own. That’s helpful. Or you could copy and paste the ten points and massively output a bunch of generic content and that’s not a great thing.
You don't always know if it’s a great thing or a bad thing - sometimes you need to just lean in and see what happens. There have been moments on the web when you look back - some people said consumers will never buy online, or do this or that - trying to shoehorn in older analogue ways into the new ways.
You need the awareness that your mind might just be based on the world you've lived, rather than the one that's on the way. You've got this massive mental capacity for creativity now that didn't exist before.
SB: You've known me for a long time. I'm generally an optimist by nature. I do fully acknowledge that this doesn't necessarily end in sunshine and rainbows, but I think there are at least plausible reasons why this could be a net gain.
It's gonna be bumpy no matter what because, if you played it out over a long enough time horizon, I suspect it would be almost universally a huge lift up to what humanity can do.
But the fact that it's happening in such a compressed time frame, there's no historical precedent for that. It's just gonna be really bumpy.
I wrote a post a few weeks ago. I’ve talked about Martec’s law - I'm heading towards changing this quickly, but technology changes so much more quickly. I've always framed that as a problem. Organisations generally change more slowly but we need to change faster. How do we keep up with this technology?
It occurred to me a couple of weeks ago that given how disruptive this AI change is going to be, actually having slower organisational change being a bit of a limiter on this could actually be a really good thing, because I think we're going to need some limiting to figure out some of these rules,
“Will AI destroy the job, or will AI augment the job? There can be a debate between those two camps, but ignoring AI entirely is not gonna be a winning strategy."
CD: Or will it just create mass disruption where you're either catching the wave or absolutely being wiped out? And you're just so slow that you get the incumbents to absolutely accelerate past you - those that have embraced this and started to understand the capabilities of AI. This is just an incredibly smart, fast brain and workflow.
So if I just take that and Microsoft it and I ask: what are my workflows? My business today with those capabilities?
Tom Goodwin asked a good question on our transformation course: If you started your business today, what would it look like? That's with these capabilities - forget your ways of working, your governance and process and documentation and how you train people.
It's interesting to think about those (organisations) that are not ready to take the leap because they want to wait to see what plays out. Will the distance between them and those that chose to take the leap be too great in a quarter? In a year?
Because of the pace at which AI is moving, if the cash moves to different companies with new capabilities very quickly, and you're not capable of adjusting to change quickly enough, are you just going to really suffer?
SB: I think that's true. Will AI destroy the job, or will AI augment the job? There can be a debate between those two camps, but ignoring AI entirely is not gonna be a winning strategy.
It seems like we have no choice. This thing has been unleashed, it’s not going back in the model. And so all we can do is lean into it and learn and adapt. I don't say that's an easy thing.
“If you can explore AI without being caught up with this new shiny thing, and then apply it to your use cases, this is the AI of the possible.”
CD: I can imagine there are quite a few companies that get lost somewhere between waiting to see what happens or doing something AI, but completely missing any thinking.
It's just the age-old shiny new object thing, it's press release stuff, it was the metaverse, Web3, IoT, and now this week it’s AI - that’s not the approach to take.
I think one of the parts that's missing at the moment is good helpful use cases that inject AI and AI just does them incredibly well.
Some of those are not yet known, but it's almost about forcing the word AI to not be used in a sentence and just getting back to age-old technology - what's your business use case and what problem are you trying to solve?
The difficult thing with AI for me is that people are asking what it can do, whereas traditionally people would start at a different point, such as wanting to sell online and being more focused on the use case. If you can explore AI without being caught up with this new shiny thing, and then apply it to your use cases this is the AI of the possible.
This was the last of a three-part interview with Scott Brinker. See parts one and two here: