How Leaders Are Using AI to Transform Business
As the capabilities of artificial intelligence grow, leaders need to be purposeful and thoughtful about how they deploy it, understanding that AI is at its best when it’s teamed with human intelligence.
Artificial intelligence surrounds our everyday lives, from sometimes-helpful chatbots to real-time maps, from labour-saving accounting apps to algorithms that keep us glued to social media sites.
While AI has become omnipotent, many businesses are yet to take a deliberate step to unlock its specific benefits. The debate over whether the natural language processing model ChatGPT is an AI superhero or anti-hero will go on but most agree it’s a game changer. The super chatbot is not the only show in town: it’s estimated that AI will be worth $22.17 trillion to the global economy by 2030.
“Australia has incredibly strong capability in the area of AI,” says Stela Solar, director of Australia’s National Artificial Intelligence Centre (NAIC). “We lead the world in field robotics, computer vision, Quantum AI, remote operations and responsible AI. Stanford University’s Global AI Vibrancy Tool ranks us fifth in the world for AI research and development or third if you measure it per capita. That’s impressive.”
Not so fast on the gloating, though. “On the flip side, our economic impact ranking for AI is 11th in the world,” adds Solar. “AI is transforming all industries and there’s a lot more opportunity for our commercial sector to benefit from Australia’s great capability.”
In 2021, the federal government’s Artificial Intelligence Action Plan found that 63 per cent of Australian businesses were having “difficulty in knowing where to start when implementing AI technologies”. It stumped up $53.8 million to establish NAIC – which sits within CSIRO’s Data61, the data and digital specialist arm of Australia's national science agency.
“Large organisations are adopting AI relatively fast and there is a lot more guidance for them, anchored on things like investing in your own data science and developer teams that you enable with AI skills,” says Solar. These big firms have the resources to fund those teams and engage consultants to help them advance their AI transformation. “Small and medium organisations are not adopting AI as fast.” Without the in-house specialist teams of big organisations, AI integration for SMEs is mostly off-the-shelf products, such as solutions for accounting, HR management, customer relationship management (CRM) and those that enable robotic process automation (RPA).
Often, the AI component in these solutions is optional, Solar adds. “SMEs need to know what their options are so that they can activate AI services and experience the benefits. The irony is that AI holds the most opportunity for SMEs, because AI can tackle scale and volume amazingly well. It can navigate vast volumes of data, which maybe SMEs don’t have the resources to do, and it can take action at scale, which smaller teams can’t do. AI is a tremendous untapped opportunity for SMEs to step up and compete with some of the larger players in the market.”
NAIC is charged with helping to make that happen, as well as connecting the Australian AI ecosystem and addressing concerns around privacy and bias by promoting responsible AI. “Our mission is to accelerate positive AI adoption and innovation that benefits business and community,” says Solar.
In March, NAIC released its AI Ecosystem Momentum report, which found that Australian organisations need an average of four partners in order to succeed with an AI project (28 per cent needed to find six partners). To help businesses navigate their hunt for the right expertise, NAIC is building an AI Ecosystem Discoverability Portal, a free online directory to help connect Australian AI innovators with customers.
It also launched the Responsible AI Network, “a network of knowledge partners to demystify what responsible AI means for organisations”, says Solar.
She’s passionate about Australia taking a leading role. “AI models are built on historical data, with injustices, biases and gaps in representation. Leaders have an opportunity to intercept the way that AI models are developed by injecting diverse skills, talent and perspective. Leaders need to be empowered to critically think through this transformation, to ask the why and shape the path that the organisation takes towards a vision of doing better in the future, rather than just replicating biases of the past.”
The legal firm: Lander & Rogers
“Lawyers are trained to look at a problem and think of all the things that can go wrong,” says Genevieve Collins, chief executive partner at Lander & Rogers. “Startups are the opposite – they look at something and think of all the things that can be done, with a lens of opportunity rather than risk. Exposing our lawyers to that is a good balance.”
They’re not doing it by halves at Landers, a 77-year-old independent Australian firm with more than 600 staff. “Since I took over as managing partner in mid-2018, we’ve been running an innovation agenda,” says Collins, “and incorporating AI technologies into projects with clients.”
The firm is using AI in contract review projects and due diligence, e-discovery in litigation and research. Off-the-shelf AI products let them run tasks such as clause comparisons and consistency checks across large volumes of documents. “In the big law firms back in the day, young lawyers would be down in a dungeon reviewing boxes of documents for 12 months – they’d leave and say they hate the law and wish they’d never done it. It was terrible.”
AI has been instrumental in removing this infamous drudge work from the profession and making space for the high-level critical thinking skills taught in law degrees. “It’s better for clients and better for lawyers because they’re moving away from the low-level review work and there are so many time and cost efficiencies,” says Collins. “It’s definitely better for lawyers from a wellness point of view.”
In 2019, Landers upped the ante by opening its LawTech Hub, at first with physical spaces in Sydney and Melbourne, moving to virtual in 2020. “It’s our pro bono contribution to the legal tech industry,” says Collins. “We don’t take equity; teams get six months of access to our lawyers and clients to test their products.” Each year, there’s an intake of six startups or scale-ups and Collins says Landers is now emphasising those built with AI.
The LawTech Hub gives teams access to basic legal advice and sessions on IP but the value is in the testing ground. “They want to test their products in real-life situations. They might find they are solving problems that might not really exist in a law firm so they can then identify the real problem and redeploy and develop a solution for that.”
There are already more than 20 alumni, including Halisok, which uses machine learning to extract and organise unstructured data from files, and DraftWise, which collects a firm’s collective intelligence to deploy the best language when drafting contracts.
AI has brought in a new way of working, and Collins says now ChatGPT is leading a shift in business that’s as significant as the arrival of the internet. “Clients don’t necessarily have access to those paid AI platforms or the embryonic startups,” she says. “The key difference with ChatGPT is its accessibility. It’s really exciting and – within our governance guidelines around privacy and transparency – we’re encouraging our lawyers to play around with it to find best use cases for it.”
She uses it herself to write the first draft of an all-staff email, Landers legal teams are using it to draft simple documents and the tech team is using it to write basic code. “I’m fascinated that some organisations are banning ChatGPT,” says Collins. “That’s just putting your head in the sand. First we had calculators then Google and this is the next iteration. It’s early days – there are clearly limitations and for lawyers a whole host of potential risks – but it’s here to stay and we have to get on board and work with it to progress it in a reasonable way.”
The big bank: Commonwealth Bank
You would expect a bank to be using the latest AI-based safety protocols to keep its customers safe from cyber attacks and Commonwealth Bank is no exception. As those essential protection measures largely play out in the background, more is needed to meet the general marketing imperative to “surprise and delight” customers.
Armed with rich financial data about their customers, CommBank has come up with some novel ways to use that information in conjunction with AI. “It’s playing a central role in building more personal connections with our customers, to really understand who they are,” says Dr Andrew McMullen, the organisation’s chief data & analytics officer.
In 2019, it launched Benefits finder, a tool in the online banking platform or app for individual and business customers to search for grants, concessions and rebates they might not know about. “There are hundreds of benefits available from state and federal governments but many Australians don’t know about them or that they’re eligible for them,” says McMullen.
“We’ve used AI to connect all of those benefits and make them available to customers we believe would be eligible. Through that capability, we’ve helped our customers get access to $1 billion dollars worth of benefits since we launched.”
His favourite story about Benefits finder is a woman who came into a branch to apply for a personal loan to replace her broken-down fridge. “One of the branch staff sat down with her and Benefits finder showed them that if she replaced her fridge with a more energy-efficient model, the government would give her 50 per cent off. They found some other benefits she was eligible for and she left with funds to buy the fridge without borrowing any money.”
In 2021, CommBank invested $134 million in leading AI cloud company H2O.ai to further ramp up its in-house AI development capabilities. The bank’s 2023 graduate intake also included 80 data science specialists. “We’re helping universities to think about providing courses and qualifications that are specifically focused on AI and machine learning,” says McMullen.
The SME: VetShop Australia
When Steven Perissinotto and his veterinarian brother, Mark, founded their online pet supplies business VetShop Australia in 1999, Microsoft Excel wasn’t quite 15 years old. Based on Queensland’s Sunshine Coast, their 22-strong team started using ChatGPT last year and he sees parallels. “I’m old enough to remember business before Excel,” says Perissinotto. “When it was new, some people were better at it than others. You’d ask someone how they got to be good at Excel and they’d say, ‘I just taught myself, I wasn’t scared.’ And that’s the stage we’re at now. We have a Slack channel and we encourage people to share ‘What I used ChatGPT for today’.”
ChatGPT training courses are springing up but Perissinotto finds there’s nothing like being your own incubator – and classroom. “A few of us were using ChatGPT in October and by late November, we said, ‘Hey, everyone should be using this.’ We found that some staff get better results than others so we’re training each other, saying, ‘Queries like this work better than queries like that.’”
Perissinotto says there also needs to be time to shake out the sillies. “You have to move staff through the novelty of ChatGPT to actually using it and understanding where they can use it.” When a friend showed him a bad haiku created by the giant chatbot, Perissonotto likened it to “when we keyed in ‘boobies’ on our calculators. Instead, you can ask ChatGPT for three ways to improve your haiku or where you can publish haiku online.”
He’s not a grinch, though. “Ask ChatGPT to write a press release and make it sound like Snoop Dogg – have 20 minutes of that for a bit of fun! Then ask ChatGPT for three ways to improve the release or give it our 1000-word article on fleas and ask it for a summary in bullet points or for the best meta tags to use for that article. We’re still all helping each other find the best ways to use it and I’m seeing a lot of hot tips on LinkedIn.”
He says ChatGPT can be a good unblocker. “Most people aren’t wordsmiths and they might be a bit apprehensive about having to email someone. ChatGPT is a great tool for that. It’s replaced ‘just Google it’ for us because it gives you a significantly better starting point.”
ChatGPT is also helping out with VetShop’s business analytics. “We can ask it in normal English, ‘What formula should I run on my database which will allow me to establish whether it’s more profitable to have a free-shipping break at $99 rather than $69?’ It gives us the formula and then we run it on the database.”
Perissinotto is pragmatic about the recommendations from the off-the-shelf products running analytics on his company’s data. “The results are only as good as the quality of the data you put in,” he says, estimating his dataset goes back to 2015. “We have to be careful that we don’t spend a fortune changing how our data is structured to get what is probably going to be an imperfect result anyway.
“AI is just a tool and you need to work out the best way to use it. We don’t believe in changing our systems to match the tool – we believe in using the tool in a way that matches our systems.”
