Software 2.0 for Martech - an account from MAU Vegas 2023

I just left MAU Vegas 2023, headed back home to Singapore. And I'm sitting at the Las Vegas airport, feeling stunned. ?? ??

The conference was 2 days of incredibly smart and experienced marketers and mobile growth professionals giving talks and having energetic conversations. And yet...

And yet, the new reality of AI has barely penetrated the mobile hive mind. There were certainly people talking about AI. But it showed up in basically no Talks, and pretty much none of the vendor booths or vendor material. No practical planning, no case studies, not much more than a vague awareness that “AI is coming”.?

And frankly, the community isn't really at fault here. It simply isn't most members’ specialty nor is it in the general marketer capability set. But whether or not it’s their fault, the change is going to be a lot more painful than it should be, because the adoption of AI is going to be a lot more variable and uneven than it should be.

All of lifecycle marketing is still stuck implementing Rule Engines the profession calls "User Journeys". Meanwhile Performance Marketing and mobile Adtech is a little bit ahead, but still trying to solve complex Ad allocation and optimization problems with logistic regression. Don't get me wrong, I love me some logistic regression, but it just isn't going to cut it.

Back in 2017, the computer scientist and engineer Andrej Karpathy wrote a prescient blog post titled "Software 2.0". That post influenced a lot of how Aampe was designed when Sami Abboud and Schaun Wheeler and I started the company in 2020, and we can use the post now to foresee what's coming down the pike, very soon, for Martech:

Here’s an excerpt from Karpathy’s post:

"The “classical stack” of Software 1.0 is what we’re all familiar with...It consists of explicit instructions to the computer written by a programmer. By writing each line of code, the programmer identifies a specific point in program space with some desirable behavior."

CRM and Marketing Tech exactly reflects this description. The software provides a UI/UX that enables the marketer to write explicit instructions - the marketer identifies a specific point in an app flow or 'user journey' to prompt some automated marketing action.

Here’s what Software 1.0 Martech looks like:

Karpathy again:

"In contrast, Software 2.0 is written in much more abstract, human unfriendly language, such as the weights of a neural network. No human is involved in writing this code because there are a lot of weights (typical networks might have millions), and coding directly in weights is kind of hard."

Here’s what Software 2.0 Martech looks like:

This is how Aampe works. Aampe customers don't have to instruct Aampe how and when to send a message, they simply provide general guidelines and then Aampe optimizes within those guidelines.

Karpathy:

"Instead, our approach is to specify some goal on the behavior of a desirable program (e.g., “satisfy a dataset of input output pairs of examples”, or “win a game of Go”), write a rough skeleton of the code (i.e. a neural net architecture) that identifies a subset of program space to search, and use the computational resources at our disposal to search this space for a program that works."

Again - this is how Aampe operates. Aampe takes an app eventstream and provides a rough skeleton of a 'treatment or action space' within which a message or ad can be delivered. We let the app marketer specify some set of goals for the app user, and then Aampe applies computational resources to search the space for a treatment/Ad/notification that works.

As Karpathy explains:

“It turns out that a large portion of real-world problems have the property that it is significantly easier to collect the data (or more generally, identify a desirable behavior) than to explicitly write the program. Because of this and many other benefits of Software 2.0 programs that I will go into below, we are witnessing a massive transition across the industry where of a lot of 1.0 code is being ported into 2.0 code. Software (1.0) is eating the world, and now AI (Software 2.0) is eating software.”

Marketing is precisely a real-world problem that has the property that it is so much easier to collect the data (the app is already doing that) than to explicitly write “the program”; in the Martech case, the campaign or user journey. And that's why Martech isn't safe from the massive transition happening across many other areas of software as well.

Ok - maybe all of the above has gone right over your head. I hear this pretty frequently when I’m talking to marketers: “I’m a marketer, I don’t understand what you’re talking about.”?

Well, I think if you really try, you can understand everything I’ve written above - you’re way smarter than you protest to the data scientist you meet at a conference.?You can't afford to not understand what is happening.

But I’ll cut to the chase and put it in terms of impact and ROI, terms you’ll definitely grasp:

Here’s Karpathy again:

“Why should we prefer to port complex programs into Software 2.0? Clearly, one easy answer is that they work better in practice. However, there are a lot of other convenient reasons to prefer this stack.”?

I’m actually going to cut Andrej off here, because I don’t think my audience for this post will care about all the technical reasons (i’ll write another post for the data scientists and engineers working in Martech). If you do happen to care, definitely read the rest of Karpathy’s post. But the first, easy, answer should be sufficient:

Software 2.0 Works Better In Practice.?

So now the question that has me wincing in sympathy for the pain ahead:?

Which mobile app marketing teams are going to recognize the future - no, in fact, the Present, because it's already here? Which teams will implement a Software 2.0 Marketing Tech stack??

Because when your competition starts using tech that works a lot better in practice, you’re going to lose pretty darn fast.?

The examples are already proliferating. Take the education app Chegg Inc. According to the Financial Times, “Chegg is deemed by Wall Street to be worth just a third of what it was when ChatGPT was launched late last year.” The company’s value dropped more than 40% after admitting that ChatGPT is killing its business.

And that’s what I’m stunned about after MAU Vegas 2023. I just saw a community where awareness of what I’ve written in this post is generally low and massively unevenly distributed.

Some teams are going to move quickly to adopt Martech (Software) 2.0. Some are just going to continue running their standard playbooks. The pain those slower late adopters are going to feel is sorrowful.

It’s a glaring bright dawn of a brand new day in Mobile Growth World. I wish everyone the best!?