3 Assumptions to Consider When Thinking About AI in Law and Accounting

December 20, 2018
3 Assumptions to Consider When Thinking About AI in Law and Accounting

A light…or dark look at what lies ahead for AI in law and accounting as we come into the holiday season.

If you’re a lawyer or accountant, you’ve probably skimmed loads of articles about the impending AI revolution. Robots are going to take your jobs, etc. You’ve also read articles suggesting that it’s not so bad. “AI will free your time from drudgery, so you can do better work.”

In this post, mostly for fun, we step back at look at AI in law and accounting without the distortion of professional optimism. After all, it’s mostly management consultants who write the optimistic pieces (they’re hardly likely to say “give up everyone!” That won’t get them paid!). Here, we take a long-term view, drawing on an interesting TED Talk.

What credibility do I have to talk about this topic? None. But I’ll do it anyway.

Three Assumptions which, if true, would render you redundant (eventually!)

These assumptions outlined by the philosopher and neuroscientist Sam Harris in this TED talk. Harris argues that, if these assumptions hold, we will inevitably create super-intelligent which dwarves human capacity in any cognitive endeavour. It’s just a matter of time.

While not speaking about law or accounting specifically, Harris argues that super-intelligent AI is inevitable if these assumptions hold:

  • Assumption 1: Intelligence is the product of information processing.
    Harris argues that “we have already built narrow intelligence into our machines and many of these machines are at a level of superhuman intelligence already” (e.g. just think about how good Siri or Google’s Assistant is at directions). As our brains are just made of matter (i.e. atoms), we know that matter alone can give rise to the ability to think across different domains of knowledge. Therefore, it’s possible to knit these domains of knowledge together artificially. In doing so, we can create machines that perform tasks previously reserved for human cognition. And they’ll do these tasks at super-human speed.
  • Assumption 2: We will continue to improve our intelligent machines
    Harris argues intelligence is our most valuable resource, and that we want to continue improving intelligence. We have problems that we desperately need to solve (curing diseases, alleviating poverty etc). As a result, he argues that we will continue to improve our intelligence. The question is how fast? Moore’s Law (i.e. the previously seen exponential improvements in computing power year-on-year) need not apply for us to continue to improve.
  • Assumption 3: We are not near the summit of possible intelligence
    “This is what makes our intuitions about [AI] risk so unreliable”, argues Harris. Intelligence extends much further than we can conceptualise. If we build machines that are more intelligent than we are, they will explore a completely different spectrum of intelligence. In doing so, they’ll exceed us in ways that we (quite literally) cannot imagine. The implications of AI in law and accounting are quite stark for professionals when viewed in this light.

 

Implications of AI in law and accounting

Harris’ three assumptions have clear implications for knowledge workers of all kinds. The capacity of machines to render much of our work redundant is true “by virtue of speed alone”. Even if we make machines that are no ‘smarter’ than a team of human professionals are (i.e. assuming that machines won’t learn to be more capable of coming to robust conclusions per-se, but will simply make conclusions faster than humans), electronic circuits operate 1 million times faster than biochemical ones. 1 week of such a machine operating can perform 20,000 years of human intellectual work. Depending on how you look AI in law and accounting, this is either terrible or great news for junior lawyers and accountants doing mind-numbingly boring work.

What the %$&# do we do?

Harris’ concerns are more big-picture and far-sighted than most pieces of thought in the professional services space are. Typically, most pundits outlook is a rosy one, for example:

  • AI revolution will simply eliminate drudgery and free up your time for more meaningful work!
  • There will be jobs we don’t even know of yet!

Maybe. But these are conclusions drawn by people who are paid to give solutions, not to say “everybody is screwed”.  Perhaps these concerns are so far-sighted that it’s like “worrying about over-population on Mars”, as Harris puts it in another podcast. Regardless, these assumptions provide an interesting framework to think about what role your firm can play in professional services.

To give this post at least some upshot, I’ve included the following questions you might ask yourself:

  1. What role will your firm play as AI in law and accounting continues to develop?
  2. Is there any readily available technology you can use now that will make you more competitive? Why aren’t you using it?
  3. Have you found technology effective to date in freeing up your time?

Post your thoughts below!

Our next post will be on Thursday 3 January 2019. It will resume its regular Tuesday posts thereafter.

About the Author

Ben Farrow is the Managing Director of FirmChecker consults with Beaton Research + Consulting, the leading management consultants to professional services firms in Asia-Pacific. He holds commerce and law degrees from the University of Melbourne and digital marketing certifications from Northwestern University.

Further Reading that might actually help you.

 

Recent Post