The CEO’s AI conundrum #2: navigating the innovator’s dilemma and practical steps for CEOs to take today

CEOs have an AI conundrum. AI has been the story of 2023, but it will also be the story of 2024 and far beyond. For CEOs and boards, your company’s approach to AI has the potential to define your business. What strategic AI choices should you make to build not just for now, but for your business’ future? This two-part series looks at the important long and short-term questions that CEOs are trying to answer right now. You can read part one here.

Newton’s law of innovation gravity

In 2007, at the launch of the first iPhone, the world gasped at a revolutionary technology. But, what if that technology had actually been delivered in 1992? How many extra gasps would have been generated?

In May 1992 a small team launched a handheld device that contained many core features of the 2007 iPhone.

The new, 1992 device could recognise handwriting, digitising note-taking. It held contacts and diary events. There were ‘apps’, including a mobile web browser.?

The device was designed specifically to have the power of a PC, with the number one requirement that it had to fit in your pocket, almost the exact same principle behind the iPhone.

Which innovative company was thinking like this, 15 years ahead of Apple? Why did they not break through with their product? What became of the ‘not-iPhone’?

The company behind the device was… Apple and in 1992 their Apple Newton PDA was announced to much fanfare at CES.

The device though struggled after a poor reception to the handwriting recognition and, on Steve Jobs’ return to the company in late 1997, he ceased development and production, just four years before launching the iPod.?

Apple, the ultimate innovators, had hit the innovator’s dilemma, killing a potentially significant product, in part because it had so far failed to generate a return.

The legacy of the Newton remains in many of our pockets today, but so too does a tantalising question in the tech and innovation realm. Could Apple have launched the iPhone much sooner, if only they hadn’t stepped away from their own innovation?

Welcome to Wrexham (part two)

In part one of this short series we looked at Ryan Reynolds and Rob McElhenney’s success with Wrexham FC, as told in Welcome to Wrexham, as an example of how a tortuous journey can become a positive and may well be a prerequisite for success.

It turns out that Welcome to Wrexham is useful for more than one business-relevant illustration.

There are twenty-four teams in English football’s League One and many of them have wealthy owners. There is only one of those twenty-four that are behaving like Wrexham and, not coincidentally, only one of those clubs who are currently on track for their second consecutive promotion.

Wrexham were not content with doing things how they had always been done and getting the same results. Their strategy has involved spending big, building a club rather than a team, investing in women’s football and lowering prices where others raise them.

CEOs and boards today stand at the pinch-point of both the innovator’s dilemma and Wrexham FC.

If we accept, as discussed in part one, that AI is a long-term endeavour, then CEOs will almost certainly face the innovator’s dilemma at some point along the way. If your reaction at this point is to kill your AI initiatives, then you face the real prospect of ‘doing what you have always done and getting what you have always got’ (which may well be inertia). Not only that, there is a realistic chance you will actually be going backwards.

In his seminal book Innovator’s Dilemma Clay Christensen describes three problems innovative, well-run firms face, with disruptive technologies like AI.

“First, disruptive products are simpler and cheaper; they generally promise lower margins, not greater profits. Second, disruptive technologies typically are first commercialized in emerging or insignificant markets. And third, leading firms’ most profitable customers generally don’t want, and indeed initially can’t use, products based on disruptive technologies.”

There’s probably at least something that rings true in there for your company and its approach to AI.

Christensen’s book is filled with examples of good, profitable, well-run companies who encountered ‘innovator’s dilemma’ and caved. They shelved their innovative projects and, as such, missed the next wave of innovation. Their fate was to go backwards; poorer performance, worse products, lower margins.

With everything we know about AI’s future and Christensen’s words in mind, what should CEOs and boards be doing right now? What does an AI company look like? How can you practically take the first steps? To put it simply: how can you make sure you guard against killing your company’s version of the iPhone?

What can CEOs and boards do now to make their business an AI business?

There are four truisms that are important to guide the following steps.

1. As we’ve discussed, integrating AI into your business - whether as a product, an internal tool or more broadly - is a long-term process and likely a significant change.
2. Whilst not true for all, many businesses not already involved in AI may be more traditional in nature and structure; think banks, chemicals, insurance, automotive and real estate, amongst others. These industries, whilst technologically advanced in their own ways, do not typically behave like start-ups, quickly pivoting to new innovative tech.
3. Businesses need to innovate, but CEOs and boards also have a responsibility to shareholders to maintain profitable enterprises that already exist.
4. In my experience, there are a number of steps firms need to go through to become good at AI. These are mostly sequential, not parallel. You must have a team and a great culture. The team, with their culture, must have access to great data. The team, culture and data can then build great products.?

With that in mind, firms should consider creating an AI-focused corporate venture capital (CVC) arm, a counter-intuitive option that can deliver significant upside.

A CVC option hedges against your high risk attempts to transform the existing business, whilst also creating a start-up-like space, unburdened with your institutional idiosyncrasies and unnecessary pressure. The CVC can speed up part of your AI journey, completing the sequential steps needed to achieve genuine innovation faster and more successfully.

Think of the alternative. Imagine a bank turning round tomorrow and deciding they are going to be an AI-first company. The weight of change required would be too much. A much easier way to produce impactful movement is the creation of the bank’s ‘AI arm’, a side venture with more freedom and a specific remit, reporting back to the CEO and parent company with gradual improvements and products that can be fed back to the parent company over time.

How should this venture arm operate? Here are five things that should be in the firm's remit from day one.

Invest in early stage AI firms

Early stage AI firms are the business equivalent of each of us individually investing in the stock market: there’s a lot of choice, a lot of risk and no-one can pick a ‘winner’ with any level of consistency. With that in mind, look for strategic synergies, rather than your ‘best bet’. If you can use a start-up’s product and capabilities in your new business then it’s likely that others can too. It is almost inevitable that even the best resourced firm will at some point purchase AI ‘help’. Think of all of the resources Microsoft have and yet they chose to invest and partner significantly with OpenAI, because it was the fastest path to show AI credibility.?

Build AI capabilities and an AI-first culture

Not only would it be difficult for a bank to change from traditional corporate to AI start-up behaviours and culture, it would be culturally unwise, to say the least. With a separate venture arm you have the luxury of embracing a start-up culture and giving your innovation space and time to grow.

Use your new venture to test and launch your own AI products

Built this way your new venture may even be able to build products that outperform the parent company. Remember that in Innovator’s Dilemma, Christensen’s case studies are all successful, well-run companies at the time they made decisions on emerging technologies that would impact them negatively in five or more years time. In the venture scenario you’re essentially making the decision to both continue with your existing successes and accelerate your ability to produce AI innovations, with the option for the parent company to reap the rewards from the innovation.

Create partnerships

CEOs and boards of large enterprises have a habit of talking a lot about eco system partnerships, which then take forever to be rubber stamped. Your venture company can operate much more nimbly than this. Use this to your advantage and secure strategic, disruptive partnerships to help your venture arm to move forwards at pace, then inject the benefit of these partnerships back into the parent company.

Feed back to the parent company

As we’ve said: it will be hard to change the parent company, but that doesn’t mean you should just give up. A successful venture arm, with successes under its belt, is perfectly placed to coach, teach and consult back into the parent business on strategic AI endeavours.

The CEO’s conundrum with AI at the moment is rooted in the challenge that AI success may look like it is a long way away. This can create inertia and lead to the shuttering of AI projects. Classic business thinking, like Christensen’s book, says that doing nothing (effectively opting for going backwards), should not be an option. The best time to solve your AI conundrum was yesterday. The second best time is today. That presents questions for CEOs and boards who do not have a plan for AI or ‘cannot find the business case’.?

How could you approach AI with a different mindset to make it work for your business? What’s stopping you creating a venture arm, with a specific remit to innovate with AI? Where do you want your company to be in five years time, when your competitors are five years into their AI journey??

For CEOs a more personal question also exists. As CEO you will inevitably leave your company at some point with a legacy. Do you want that legacy to be that you missed your company’s ‘iPhone moment’??