For AI Visionary Craig Macaulay, a data scientist at Phi Finney McDonald in Melbourne, Australia, artificial intelligence is not just important—it’s inevitable. In fact, he’s so excited to see where it leads that he can’t imagine retiring. Craig sees an abundant future thanks to AI, featuring incredible time savings and greater access to justice. Take a look at his tips on how to dig deeper and get more out of the technology in your day-to-day work.
Ryan Docherty: What were your interests early on and what drew you to your line of work?
Craig Macaulay: AI a natural evolution of the power of data. Thinking back to early 90s, you could see that data analytics was going to be the future of business. It did not take off as quickly as some predicted, but now it has exploded. I always knew that analytics was powerful and could improve learnings.
Even analytics has progressed from the traditional structured analytics, which can tell you the “what” of your data. But assessing unstructured data, such as emails and documents, offers the insight to tell “why”. Today, it’s clear we need to understand AI and machine learning and how it can help manage and analyse unstructured data.
Please tell us a little about your role and what you enjoy about it.
I really enjoy constantly seeing new tools and applications that are out there and being able to deploy them inside a business. There is so much more to achieve with technologies like Bert, Google’s search tool, or ChatGPT. I wish that there was more time in the day to test and implement the technology and examine how to harness and evolve it for our space.
One area I was recently able to put this into practice and leverage my background in data science was as an early user of sentiment analysis in RelativityOne. As the Relativity team fine-tuned the feature, I was able to collaborate closely and provide feedback to influence the design and use of the technology, helping them achieve high-quality outputs from the workflow. This is the sort of testing and innovation incubating that I really enjoy in this role. Overall, as a legal technologist, it’s about using the vision to see what you can do with technology to have an impact on everyday work. My role is not highly technical; the greatest skill required is imagining how to help in improving a business process and then uncovering the tools that can assist.
Any advice for those who are interested in following your career path?
Use your imagination but stay grounded. The tricky bit is not producing a new tool or application, but establishing how you can embed it into workflows and thought processes. If you can’t achieve this, it won’t be beneficial. In data science, we tend to think too much around how to do something rather than how to embed in the workflow and thought processes.
What is your vision for the future of AI? What are the biggest areas of opportunity you feel AI can have an impact on?
Traditionally, the legal industry gets bogged down in process. We spend 70 percent of our time managing processes, which leaves just 30 percent of our time to spend thinking about the best way to manage a case. This should be reversed. What would happen to litigation if, for example, AI could create all of the pleadings and complete standard tasks on its own?
Starting with a new document is the hardest part, right? Imagine if you could use an algorithm to begin, and the legal practitioner finishes it off. It would save a huge amount of time.
I think this would lead to greater access to justice and could really unclog the legal system.
What’s your advice for organizations hesitant to adopt AI?
Know that it is career limiting if you choose not to get on board. Businesses will have a finite life in this industry if they don’t adopt AI.
How can AI enthusiasts develop in this area to close an AI skills gap in your industry?
Spend a lot of time thinking about what you can do with your data. Start by finding a problem and imagining what you can do with the data involved to simplify it. A level of programming understanding helps, but you really need a level of data understanding. That enhances the ability to understand how data can be used.
The other problem we face is how to integrate AI into workflows and thought processes. The best analogy I can use is from The Hitchhiker’s Glide to the Galaxy, in which a supercomputer called Deep Thought is asked: “what’s the meaning of life, the universe, and everything?” After 7.5 million years, the computer comes up with the answer of “42.” Now, from the AI’s perspective, this may be correct. However, as AI practitioners, if we can’t contextualise the answer for users to understand it is of no benefit to us.
We need to be able to bring the user on the journey to understand how the AI has come to its answer. This also applies to other areas of AI, like algorithmic bias. Issues like this are always going to arise, but if users can understand how it has impacted the algorithm, they can apply their own judgment.
In The Hitchhiker’s Glide to the Galaxy, a supercomputer called Deep Thought is asked: “what’s the meaning of life, the universe, and everything?” After 7.5 million years, the computer comes up with the answer of “42.” From the AI’s perspective, this may be correct. However, as AI practitioners, if we can’t contextualise the answer for users to understand it is of no benefit to us.
Are you seeing an AI skills gap in your industry, and if so, what will help close it?
What we need more of is confidence, and a desire to actually pursue these things. Teams have to get out and do this first. Just get started and try it. That confidence level will be built up. I truly think this is a mindset issue and less about a skills gap.
Change your mindset and think of how technology can harness the information in your data instead of how it might threaten you.
That being said, AI practitioners do need to carefully consider how to engage users, integrate, and then embed AI in workflows and thought processes.
What is unique about adopting and implementing AI in your region? How are new legislation and data regulations in APAC inspiring or inhibiting the use of artificial intelligence in e-discovery and compliance?
An antiquated mindset hinders us; we still have practice notes about exporting MDB files during the exchange of data. That is a proprietary file format that has been deprecated by Microsoft since 2003, yet it is still embedded in document exchange practice notes.
The weaponisation of using new technology to start fights among legal practitioners is also problematic. We have cases out of the Australian jurisdictions supporting AI and technology-assisted review; how could someone still deliberately block the use of tools like predictive coding at this point?
That said, the Supreme Court of Victoria is working to ensure that technology-assisted review is used at all times where appropriate. The delaying tactics can’t work for long.