Baker McKenzie's Ben Allgrove on why law is now a 'data-aggregation business'

There is one problem, he jokes. "They might cut me out and rely on the computer."

Allgrove topped his year at University of Adelaide and left for London as a Rhodes Scholar in 2002. He also topped his Bachelor of Civil Law course at Oxford before completing a Master of Philosophy on the Legal Personality of Artificial Intellects.

He joined Baker McKenzie in 2004 and by 2010 he was a partner in its Intellectual Property, Technology and Communications Practice Group,

'An evangelist for change'

Allgrove was in Sydney talking to 10 of the firm's top clients and said he relished any chances "to be an evangelist" for change.

He reckons "there's lot of hype in the market".

"What's currently happening in the market is business optimisation, not innovation."

The next two to four years will be about "service redesign". Thereafter, he says, the progress "will be determined by how far along the machine learning products come on."

Even the advanced products "are not there yet".

"We test them and we still get better results using traditional databases and basic search.

"But once you get those expert systems that can be trained up to deliver legal knowledge ... that creates some interesting possibilities."

He expects "we will end in a world where clients are buying solutions as a service with different pricing models".

"Some of it will still be time and materials, others will be subscription, others will be pay-per-drink. You will probably have tech vendors in the package and other expertise.

"If it's employment you might have HR consultants, whether they are part of a law firm or in a JV [joint venture] with a law firm.

"We'll try all those models and see what works."

'A raft of new providers'

Allgrove's crystal ball says that in five years "there will be a raft of new providers offering quote unquote legal services".

"There will be technology vendors offering solutions direct to clients. There will be alternative legal providers with different staffing models – your Airbnb-type model for lawyers. And there will be traditional law practices.

"No one really knows where it's going to come out. What we are trying to do is apply design methodologies into the services space ... to sit down with clients and work out what those service lines might be."

For example, the firm is working with a Canadian bank on a program which will take the bank's commercial contracting largely out of the hands of its in-house legal team and also do the risk analysis.

"It's a learning system that over time will get smarter and smarter."

Allgrove says machine learning is tough to do right.

"We are using it in our trademarks practice to handle large volumes of information," he says.

"When the machine learning people came in to automate bits of that process, they were pretty cocky about the whole thing and how easy it was going to be.

"None of them have cracked it yet, and we are working with the best in the market."

Part of the problem is that the firm, which has the largest trademarks practice in the world, operates in 204 jurisdictions and uses 80 languages.

"Once you get into places like Tibet, it becomes complicated".