The atomic unit of strategy original post by Ryan McManus

Artificial intelligence will accelerate change in every dimension of the digital economy.

For over ten years we have worked with boards, investors and senior leadership across corporations, governments and the military to help them understand the growing urgency of digital transformation and the need to rethink business models, investments and culture. Every discussion included the coming rise of AI, which was always predictable as a result of the explosive growth in data volumes and cloud-based computing power, and how AI would re-platform every element of the digital economy which preceded it. Early examples, such as Google’s Sidewalk Labs, demonstrated its potential several years ago.

Still, the emergence of generative AI, arriving with unprecedented speed, has caught many leaders by surprise. Leaders need to understand how AI will change their businesses, and grasp this change much faster than many did in the previous era of digital transformation. Speed and scale win in the digital economy, and AI drives both at supersonic levels. It’s very simple: AI enables organizations to deliver what was previously impossible. How should leaders respond? And what data foundations do businesses need to prioritize to stay relevant?

The golden age of AI

AI is already mission-critical in many sectors, and has also reached the point where it can startle. That is reflected in an accelerated uptake. Netflix, which launched in 1999, took 3.5 years to achieve one million users. Instagram, launched in 2010, took 2.5 months. When ChatGPT launched in late 2022, it took just five days to hit a million users – reaching 100 million within two months.

The rapid success of ChatGPT has highlighted the democratization of inexpensive access to advanced digital capabilities. While the content and image generation capabilities of ChatGPT, Dall-E and other companies are impressive, generative AI’s impact is rapidly expanding across sectors.
Take, for example, the July 2022 announcement by AlphaFold that its AI capability had solved the previously unsolvable problem of predicting protein structures. This historic accomplishment enabled researchers at the University of Toronto and Insilico Medicine, using the Pharma.AI platform, to create a new solution for the most common form of liver cancer, and to predict the patient survival rate. Time to solution? Only 30 days. Science went from a highly-complex, insolvable problem to new solutions in a matter
of weeks.

AI enables teams to do things which were previously thought to be impossible. As similar examples rapidly proliferate across sectors, the challenge for leaders is to think differently about AI-driven competition, create a shared sense of urgency, and create new structures which allow their organization to keep pace.

Building blocks of data

Algorithms hold little value without significant data sets to power them. The lifeblood of AI is data. Yet it is often treated as tactical, rather than as a fundamental driver of competitive advantage and new capabilities. MIT Technology Review Insights and Databricks research shows that only 13% of companies are utilizing their data properly. Why are leaders ignoring this critical element of digital transformation? Gartner has identified the four levels of capability that data teams have typically focused on:

1. Descriptive What happened?
2. Diagnostic Why did this happen?
3. Predictive What might happen next?
4. Prescriptive What should we do next?

But a fifth level, enabled by generative AI, is by far the most exciting and transformative. This level asks: what can we build?

This level is only attainable with the most advanced data capabilities: organizations still struggling with rudimentary data management risk cannot compete at this level. The companies pioneering Level 5 are prospering: the median pre-money valuation of these generative AI startups has reached $90 million in 2023, up from $42.5 million in 2022, according to Pitchbook.

Industry disruption

Even Level 1 data capabilities can upend entire industries when applied at scale. In the music industry, for example, the identification of emerging artists has changed forever as the amount of data has increased. Vasja Veber, co-founder of Viberate, the music analytics platform, says: “Nowadays the first step of finding hot artists is done with analytics, which wasn’t possible before streaming media and social media. We have 700,000 artists with verified profiles and can check who are emerging as the hottest artists.”

Traditionally, only major record labels offered advances to artists, based on their evaluations of the artists’ sales potential. In return, they owned the rights to the artists’ catalogues. But with the influx of data into the industry, artists can now directly calculate the potential value of their music (the predictive level), and either self-finance or take loans based on their expected value. Companies like Concord Music, Primary Wave and Hipgnosis have even used analytics to spend billions on acquiring song catalogues, using data to determine the future value of the music.

As a result, the economics of the music industry have been changed significantly. “Before streaming, artists were monetizing by selling mechanical media,” says Veber. “Once the CD was bought, that was the only revenue the artist received. With streaming, artists make money with every play, so the lifespan of a track can now be unlimited, revenue can be unlimited. It has enabled the whole industry. It’s basically fintech meets music,” he adds. As data enabled predictive capabilities, record labels lost much of their traditional advantage.

Tesla is also increasingly heralded – not only by Elon Musk, but by investors – for its AI leadership. Tesla captures an astronomical amount of data from its fleet of electric vehicles, its energy businesses, and its advanced robotics manufacturing processes. It captures so much data that it has its own AI chips and software platform to manage its data operations. This allows Tesla to develop AI solutions which may eventually be sold as a service across manufacturing, autonomous transport, and even personalized robot assistants – rewriting business models across multiple sectors. Recognition of such growth opportunities gives Tesla a market cap multiple times larger than its competitors.

Rethinking the value of data

Explicitly valuing data allows leaders to highlight the importance of this foundational element of digital transformation to their organizations. David Bach, the chief executive and founder of Optios, the AI-powered neuro optimization company, is working to create the world’s largest database of brain data. “Data aggregation, management and optimization becomes the primary source of competitive advantage,” he explains. If you get it right, the value of data instantly soars. “Lots of companies have spent tons of money and gotten zero value. It’s because having data doesn’t do you any good, unless you figure out how to garner a competitive advantage from it,” says Bach.

Shelley Leibowitz, board director at Morgan Stanley, Elastic NV and several privately-held fintech and information security companies, also places an emphasis on using data correctly. “It’s about ‘what do we do that’s unique for our customers? What do we do better than anyone else?’ And then, ‘what information do we have that supports those things?’” she says.

Keeping up with the speed of AI

The amount of data has exploded in recent years, as has the advancement of the tools required to get the most out of that data. “The amount of data that’s available to us all today is thousands of times bigger than it was decades ago. Figuring out how to analyse it, and learn from it, gives you opportunities that were unimaginable years ago,” says Bach.

Speed is key. “If we used two-year-old or even six-month-old machine learning systems we would be so far behind the curve it would be embarrassing. If you are using technology from several years ago, you are exposing your company to a ton of risk,” he adds.

To address those concerns, companies – and specifically their chief executives – need to make data a priority at the highest level, says Bach. “Say that this is a huge strategic priority, and spend time talking to the people who work on it. Participate directly in sprint planning meetings and scrums. Have your data management leader report directly to the CEO, even if they are the least senior person in the company.”

The speed of progress in this field can seem daunting. Bach says the pace of change can “boggle the mind”, but it also comes with an upside. Keeping pace with emerging AI capabilities saves money. Bach describes the changing economics of AI: “Smaller dollars can yield larger benefits. Three or four people with current data can outperform a team of 50 with slightly older technology – AI has quintupled the productivity of software engineers.”

AI strategy and mindset

How do you ensure that your organization’s mindset is keeping pace with these changes? Shelley Leibowitz says data and intelligence
“need to become actions”. Isolated data is irrelevant, and too often businesses are organized in a way that creates silos, meaning data can’t be orchestrated effectively.

It’s also important to treat data less as a minefield that needs to be navigated, and more as the underlying structure on which to build your business model. Many boards still view data as a risk to be managed, due to cybersecurity and other issues, but that’s a limited approach, says Leibowitz: “Very rarely is data something that is talked about in the boardroom from a strategy perspective.

“Leaders don’t talk enough about changing business models and allocation of capital to real investments in things like tech and data. It’s more about strategy than risk management, a risk management context puts people in a defensive posture. What is the forum that is not just about risk management? It all comes back to capital allocation,” she adds.

Putting all this into practice can seem like a gargantuan task, but leaders can make an easy start by asking the following questions:

• What is our level of data capability?
• Which C-suite executive owns our data and AI strategy?
• How do we value our data?
• What risks and opportunities does AI pose to our value chain?
• How do we know that we are not thinking too small?
• How do we keep pace?

By addressing these areas, leaders can put data and AI at the heart of their business, but it does require an explicit strategic focus. “There are very few things that technology cannot do today,” explains Leibowitz. “The technology is all available, it’s about mindset and behaviour, and being able to envision something different which creates value for your markets.”

AI will continue to present significant challenges, not least around the ethics of its use, privacy and cybersecurity, all of which leaders will need to address. At the same time, leaders must begin to address the generational challenges represented by advancing data and AI capabilities. How does your strategy and leadership need to change when technology makes real what was previously considered impossible?

Ryan McManus is an educator for Duke Corporate Education, board director, and the founder and chief executive of Techtonic.  He also partners with us at Valize on selective engagements.  

*Originally published on Thought Sparks