Unlocking the potential of AI in your business

You’d have to be living under a rock not to have heard of the generative artificial intelligence (AI) tool ChatGPT-4, released to the public by OpenAI in late 2022.

It’s among a selection of large language models, such as Google’s Bard and Microsoft’s Bing being widely adopted the world over. And these technologies are quickly transforming the face of many industries.

So, how do business leaders know which tools to adopt? And how to build the culture to do it successfully? Where will competitive advantage come from? And how seriously should leaders be taking warnings of “profound risks to society and humanity”?

Instead of asking ChatGPT for answers, we turned to Professor Alan MacCormack, Harvard Business School Adjunct Professor of Business Administration, and member of the Digital, Data and Design (D^3) Institute at Harvard.

An expert in innovation management and new product development with a focus on digital technology design and deployment, Professor MacCormack has a long history of research, consulting and case study exploration of technology giants such as Microsoft, NASA and Amazon.

From chess to chatbots

Tracing its roots back to the invention of the modern computer, which enabled automation of data processing and decision-making tasks, Professor MacCormack adds a bit of perspective around AI by viewing it as part of a bigger picture of software advancements and just one of many tools being used by organisations.

Noting that it’s been a while since a computer first beat Russian former world champion Garry Kasparov at chess, and that there have been many other fearful “false dawns”, Professor MacCormack acknowledges that what we’re seeing now is clearly an inflection point in its development.

He explained how a confluence of factors, including the availability of enormous amounts of unstructured data, increased computing power and advancements in algorithms, are contributing to this significant turning point.

“Programming a computer to play chess is a very defined task, whereas we now have these models that can take lots of general data and make sense of it. Our current models of chatbot have passed the bar exam and the medical exam,” he said. “And that’s pretty inspiring and maybe a little bit frightening.”

The data advantage for business through AI

Among its benefits, adopting AI can unlock new revenue streams, streamline processes and leverage data insights for strategic decision-making.

According to Professor MacCormack, for most organisations, adopting AI wouldn’t necessarily involve hiring their own internal data scientists and building algorithms.

“It’s actually going to be about ‘how do we marshal our data and connect it together in a way that it’s analysable, so we can actually use it to make decisions and discover hidden patterns that we weren’t aware of?’” he said.

A critical question for businesses to ask is where competitive advantage will come from and in Professor MacCormack’s view, it’s likely to include using externally developed tools alongside internal business intelligence.

“How do we use external resources that already have these models and algorithms developed and have generated what we call API's or interfaces that allow us access that technology?

"And how should we use that with our data? A lot of our advantage will lie upon using our proprietary data,” he said.

The network effect

The deployment of AI can uncover connections between seemingly unrelated variables, to broaden business intelligence.

“What we call ‘network effects’ are at play here, which refers to when the number of pieces of data that you have, or the number of users and the number of nodes in the network increases, you get value from being able to put all of these things together, " said Professor MacCormack.

“When we think about artificial intelligence and what you can learn, these network effects are actually applied to data. Many firms have lots of data, but they're in silos. So, we need to find a way to marshal all that data and create data pools.

“The promise of artificial intelligence is, the more variables you throw in there, even if they're seemingly unrelated, you discover relationships,” he said.

Is AI relevant to all businesses?

Some business leaders, particularly of smaller organisations or from more traditional industries, may question the relevance of AI to their organisation. In Professor MacCormack’s view, AI presents some form of opportunity for every company.

“If you’re in a more traditional industry that’s not as data intensive, it’s not as though you don’t have data, so there is opportunity there.”

Not engaging leaves the business in danger of falling behind competitors, resulting in loss of market share and fewer efficiency gains.

“There are a lot of companies that are data intensive and have been using pretty sophisticated models for a long time. If you're in the finance industry or an insurance company, you're used to modelling risk. You're used to trying to understand everything you can about a consumer. So, to them this is just the next generation of decision-making technologies.

“It’s companies that haven't yet made that leap, and they're not really deploying algorithms and data and analytics currently, that are really kind of faced with the question of ‘what do we do at this point?’ And doing nothing is really not an option.”

Absorptive capacity

Using the term 'absorptive capacity' from academic literature, Professor MacCormack referred to a business’ ability to “understand technologies and new developments outside of your field that may impact you.”

Determining which aspects to automate or replace is crucial, along with analysing whether the skills required to make these decisions and effectively use AI tools already exist within the business, or need to be acquired, or developed.

 

For Professor MacCormack, organisations need people who can champion research into AI and help leaders assess the potential for using it for competitive advantage.

“In order to be a ‘wise consumer’ of AI, you need to have some people internally who know a little bit about it,” he said.

Turning risk into opportunity

For some organisations, it’s hard to know whether modern technologies will pose a threat to the business or give it a boost.

Professor MacCormack points out that it can be both.

“I always think that risk and opportunity are actually just opposite sides of the same coin. A risk gets turned into an opportunity if you spot it early and leverage it in a way that potentially enables you to create value.

“If you saw the internet five years ahead of everyone else, it wasn't really a risk to your business. It was really an opportunity. And I think that's probably true here if you get started early.

“And getting started doesn't mean that you have to spend a tonne of money. But accept that this is coming and this is going to affect every business, no matter how big or small, how advanced, or traditional.”

When it comes to action and preparing the organisation’s employees, Professor MacCormack emphasised the value of fostering a culture of continuous learning and adaptability within organisations. Companies should focus on both retraining and recruiting to make the most of new potential applications.

“For many firms, they may not have people inside who are familiar with these areas… culturally this is going to be a little bit of a shift,” he said.

Actions for embracing the AI opportunities

For successful AI engagement, Professor MacCormack identified the importance of building both an “experimental and agile mindset” and engaging in these activities:

1. Create champions within the organisation

Ask a few people inside the organisation to make it their job to understand AI as part of their role and act in an advisory capacity on possible technologies to adopt and potential impact.

“That might mean allocating a few people and saying, ‘Okay, your task now is to report back within 90 days and run a companywide seminar on artificial intelligence and what it means for us.’

“Just that effort alone will be giving people license to go off and explore and experiment.”

2. Identify opportunities and apply resources

It’s important that the exploration of AI within the business doesn’t end with the report-back session. It’s what comes next that separates the ‘thinkers’ from the ‘doers’.

That can be allocating budget, upskilling existing staff, or recruiting new talent to identify opportunities and bring them to life. These create momentum and early lessons for the business.

Having access to people who understand the data models required for the organisation’s future is a key asset.

3. Get your data in order

Professor MacCormack predicts that, when looking at their data in order to explore potential AI opportunities, most organisations will find that their data is disorganised.

“Having data everywhere means you’ve got data nowhere,” he said.

“If unique advantage in the future is going to come from everything that we can learn from our data, and how we can make better decisions and equip our managers to make better decisions, we have to get our house in order.”

He suggests that you start to categorise what you know about important segments of your business and if you find that the data is insufficient, then augment it from additional sources.

"And so 'how do we get new data to combine with the data we know and then leverage all of that in a way that helps us make better decisions?'”, he said.

4. Test and learn

Understanding how your customers and other stakeholders will respond to a new model or service requires testing. However, Professor MacCormack has seen it play out many times, where leaders believe that because their people have studied something and there has been some discussion around strategy, they can predict with certainty how it will all unfold.

“Firms tend to think they know stuff when they haven't necessarily done anything. We sometimes refer to this as the ‘knowing-doing gap’.

“But unless you do something, absolutely, you don't really know,” he forewarns.

“Find early opportunities to beta test, or nowadays the term that's often used is ‘minimum viable products, or minimum viable services’.”

He also recommends that deployment is not so much about making money, it is about learning.

“And then there's a feedback loop…which opportunities are really big here, but hard to execute in these other areas. Or that the technology's not quite ready for ‘primetime’.”

Balancing curiosity and concern

But should we be engaging at all? Like anything new on a global level, AI advancement is understandably evoking strong and mixed emotions.

On one hand, it's being hailed as the new superhero in its ability to save our lives through health technology advancements. We’re hearing how it will improve job satisfaction by automating mundane tasks, freeing our time for more purposeful pursuits. 

On the other hand, we’re being warned it threatens to profoundly damage our lives, perhaps our very existence by issues such as displacing labour markets, its potential for rampant plagiarism, a tsunami of misinformation and the faster, cheaper production of chemical, cyber and other weaponry.

An open letter to AI labs calling for a 6-month pause on “the riskiest and most resource intensive AI experimentation” so safety protocols can catch up, has been signed by researchers and leaders including Elon Musk and Steve Wozniak, with warnings also from OpenAI CEO Sam Altman and ‘Godfather of AI’ computer scientist Geoffrey Hinton.

Given the global nature of AI research and development, Professor MacCormack highlighted the challenge for any single country to unilaterally implement a pause or enforce regulation.

Keen to point out that the signatories to these statements are closer to the specifics of the threat of AI to humanity than he is, Professor MacCormack supports the need for international cooperation and collaboration in order to safeguard future developments in AI.

He suggests differentiating between research and deployment, with extra caution and scrutiny applied to the latter.

“We have to think about this in a global context versus a single country context. We have to separate the different elements of the AI process,” he said.

The future is now

Despite concerns, with the emergence of new artificial intelligence models capable of processing vast amounts of data and providing “compelling answers to complex questions” - one thing is for certain for Professor MacCormack – AI is here to stay.

He urges leaders to engage with care. “You have to be a considered consumer of the technologies that are out there,” he said.

“Avoid at all costs, the ‘Big Bang’ syndrome, which is we're going to put a team of 1000 people on this and in five years, we're going to release our thing. And before you know it, you'll find out that there was a lot of stuff going on in other companies and you just couldn't see it.”

Business agility involves quickly and inexpensively experimenting with new ideas and accepting failures as part of the process towards success. Professor MacCormack acknowledges how difficult this can be, especially for larger organisations.

“What happens is you end up having these routines and processes and resource allocation mechanisms that act as a glue or a cement around your existing business model and around your existing technologies.”

A critical take-away from Professor MacCormack

“Artificial intelligence is already here. And it is an important part of your business right now. Even if you don't recognise it. So not having a strategy is having a strategy actually. Ignorance is a strategy - one that is setting you up to fail in the future.”

Professor MacCormack co-facilitated the Senior Executive Forum 2023 where he helped WA leaders improve their organisational agility and foster a culture of innovation.

One of the things they talked about at the forum is ‘how do I unstick some of the internal workings of an organisation to become a little bit more agile and to bring about a culture that is more experimental?’”

Professor MacCormack explored how organisations can embrace experimentation, making it acceptable to try new things, even if they fail.

“The only way you'll ever get to the future is by having several failed experiments,” he said.

Successful strategy, product or service model transitions are not without their challenges. Professor MacCormack explored what leaders need to do to carefully manage the decline of outdated business models, while simultaneously embracing new opportunities and acquiring the necessary skills and talent.

Where to from here?

Join Professor Ethan Bernstein and Professor Tiona Zuzul to challenge and expand your thinking and unpack the complexities of the issues interrogated through Harvard Business School case studies. Book now