Driving business outcomes in "The year of AI"

This will be a transformative year for AI based technology

Based purely on the volumes of strange & hilarious AI generated art (see?DALL.E 2), or the surprisingly human answers coming from?ChatGPT, this will be the year of AI based technology. Despite the perception of major breakthroughs, these examples of generative AI are actually incremental improvements of well established algorithms which have been around for several years. Their developers are getting better and better at applying the appropriate training datasets, and making their algorithms accessible in different ways that enable more creative interactions with the underlying models. While there may not be many direct applications for these memes (I’ve pasted my favourite at the bottom of this blog), it’s clear that AI has once again captured the imagination of many people, and will consequently generate a lot of excitement.

Beyond that hype, there is also a significant volume of well-considered, thoughtful investment going into everything from Venture Capital (Aussie VCs ready for the next tech boom: Generative AI) through to the continual improvement of the Platform-as-a-Service (PaaS) offerings from the big three hyper-scalers (Microsoft, Google & Amazon).

When I started working (in a Price Optimisation startup), we had to code our own Machine Learning, Simulation and Optimisation software from scratch, in a combination of C# and C++. The core application & asset which differentiated our business, took several dozen person-years to build. If we were to develop these applications today, PaaS platforms would help us build far more capable and user friendly applications, at a fraction of the time and cost. To give a more specific example, my final year project at university, involved building a robotic vision system, which did (low quality) recognition of objects as circles, squares or triangles. The many hours we spent coding up the algorithms, would now be replaced by a handful of classification API calls to?Google Tensorflow?or?Amazon Rekognition, which would also do the job many orders of magnitude better than Abi Virandra and my coding skills.

The business models of these hyperscalers is underpinned by selling more storage and computation. By implication, they continue to invest billions of dollars each year, to continually build up their platforms, and create new, easily accessible AI capabilities, which dramatically reduce the cost of innovation for businesses. As a consequence, it’s clear to see how the age of AI & Cloud technologies are combining, to continually drive innovation (through lower cost), disruption (through new capabilities) and increased competition, across all sectors of the economy.

AI is a commodity which enables differentiation – it is not a differentiator itself

While there are a vast number of algorithms and applications in the AI universe, they group into a smaller handful of business capabilities, with very little underlying differentiation. Major breakthroughs in model training and prediction are primarily in the public domain, and replicated amongst a large number of software applications (both on-premise and in the cloud). In addition, many models are openly accessible and pre-built (like the image recognition ones mentioned above), and access to data is better than it’s ever been before. This is especially true when it comes to the vast range of?easily accessible external data, and the ongoing trend to empower people with the ability to share their own data, most notably through initiatives like?Open Banking. Easy access to these data reduces the incumbency advantage of existing organisations.

As a result, the underlying AI technologies are being rapidly commoditised, and made accessible without significant capital investment requirements. Consequently, these algorithms (on their own) won’t create any lasting competitive advantage.

AI will create competitive advantages, but only through business model innovation

Being a strategy consultant, you’re often asked for a list of “the best use cases and benchmarks” in a topic, and increasingly, for the application of AI in business. The focus on use case catalogues, ignores the fact that by the time these use cases are in the public domain, they’re so dated and commonplace, that they are, at best, examples of Level 1 / Advanced Digitisation in the framework below (from?The CEO role in digital transformation). Amazon and Netflix were pioneers of personalised marketing – these capabilities are now part of almost any retailer with a large product range. Use case catalogues absolutely have a place in business strategy & planning processes, but their primary role is to highlight what is or will soon become a table stakes capability within a given sector.

Depending on the current state of your organisation or business function, you may need to make significant investments, in order to meet these table stakes capabilities. To move beyond that, and have a sustainable, ongoing business, the underlying business model & way in which you create value, needs to evolve as well. To that end, the organisations which will thrive in this era of AI, are the ones which are applying these technologies, to enter new markets with dramatically improved economics, create new products and services that weren’t previously possible, or to build new and disruptive business models, which capture value in different ways.

Intuitively, this may sound like something that only a handful of extremely well capitalised, global mega-corporations can do. However, research Deloitte developed with The University of Melbourne, highlighted over a dozen positive examples across all sectors and size of organisation. That’s the beauty of commoditised & accessible technology, and is the ultimate reason of why you’ll only create value by integrated technology into the business, rather than hoping that technology creates an advantage all on its own.

Thank you for reading, and if you’ve made it this far, I wanted to show you my favourite ChatGPT meme, courtesy of Anna Clive searching for her top ten strategy tips off ChatGPT:?“The best strategy is to be strong. But if you cannot be strong be clever and deceptive.” -Sun Tzu

This post was also made on www.bizmodeltransformation.com, where these topics are explored in much further detail, together with the research that was conducted with The University of Melbourne.