Innovation paradox
|
JC pontificates about technology
An occasional series. This article derives from a 20 year meditation on a conversation with C.E.M.C., for whom many thanks and great respect.
|
Why do reg tech solutions promise so much but deliver so little? This is the Innovation paradox. Is it a paradox, though?
"We don't pay lawyers to type, son"
Classic example: computers and the law. In 1975, when you wanted to edit a legal contract during the negotiation that would mean retyping the entire page. Hence, legal comments in a negotiation were necessarily bounded by the effort and time of recreating the document. There was an art to saying something once, clearly and precisely. Since editing was wasteful exercise, superficial amendment was not, for the avoidance of doubt, the apparently[1] costless frippery it is today.
Things weren’t so bad in 1975. There was a natural limit on legal wrangling. The physical cost partly negated the anal paradox.
By 1995 lawyers had computers on their desks, and the traditional refrain[2] "we don't pay lawyers to type, son" was beginning to lose its force.
Suddenly, it was easy to re-generate documents, to insert new clauses, to futz around with words. Sending documents was free and instantaneous. Far from accelerating negotiations and enhancing productivity this gave free licence to pedantry. Negotiations quickly became convoluted and elongated. You argued about trifles because you could. It also lowered the bar: certain classes of agreement which previously could not justify their own existence, let alone legal negotiation, could now be thrashed out and argued about.
I have no data for this — where would you get them? — but I am certain the number, length and textual density of legal contracts exploded after 1990.
Many of the artefacts of the analogue era of negotiation — the gremlins and hair-balls you would expect technology to remove — persist to this day. We still have side letters. We still have separate amendment agreements. We still have solemny write this page is intentionally left blank. We still have this clause is reserved, because no-one’s managed to figure out Word’s automatic numbering system. Not only has regtech failed to remove expected complexities, it has created entirely new ones.
Why is this? It is a function of the incentives at play. Lawyers and negotiators are remunerated by time taken. They are rewarded for the complexity and sophistication of their analysis. Lawyers don’t want to simplify. Lawyers don’t want to truncate. That isn’t in their nature. It is contrary to their nature. This is not what lawyers will use technology for. Lawyers will use technology to find new complexities. To eliminate further risks. To descend closer to the fractal shore of risk that they see it as their sacred quest to police. But that shore is fractal. However close you get to it, the risks remain.
We lawyers use technology to indulge ourselves.
Technology has been brilliantly efficient in allowing lawyers to showcase the sophistication and complexity of their syntax.
There is a serious point here for people (like me) who argue that technology implementations should be driven as far as possible by users at the coalface. And that is to bear in mind that the interests of users at the coalface are not necessarily aligned with those of the organisation for which they are working.
Natural language processing
A more recent example is that of natural language processing. There is a well-known and widely celebrated example of an application which cuts out legal work by performing a preliminary review of a standard agreement such as a confidentiality agreement against a preconfigured playbook of policies. The idea is triage. The machine will scan the agreement and pick up the major points against the firm's policy and highlight these for the lawyer who can then quickly deal with the points and respond to the negotiation. The application proudly points to a comparison of their software against human equivalents in picking up policy points in a sample of agreements.
But it get the triage backwards. Rather than having the lawyer pick up the major points (the high value work) and then employing the AI to process and finalize the detail, it is the AI which picks up the major points and tasks the lawyer with completing the clerical work. For the process to be productive the lawyer must rely on the AI to have identified all salient points. Otherwise, the lawyer must read the agreement in full as a sense check. In practice, natural language processing is not sophisticated enough to allow this level of comfort, nonetheless lawyers are encouraged to trust it. Hence a buried risk.
Furthermore the reality is that many of the policy points in the playbook will be non-essential "perfect world" recommendations ("nice to have s") which an experienced negotiator will quickly be able to wave through in most circumstances.
But this software is designed to facilitate "rightsourcing" the negotiation to cheaper (ergo less experienced) negotiators who will rely on the playbook as guidance, will not have the experience to make a commercial judgement unaided and will therefore be obliged either to escalate, or to engage on a slew of nice to have but bottom-line unnecessary negotiation points with the counterparty. Neither are good outcomes. Again, an example of reg tech creating waste in a process where investment in experienced human personnel would avoid it.
The basic insight here is that if a process is sufficiently low in value that experienced personnel are not justified, it should be fully automated rather than partially automated and populated by inexperienced personnel
The Jolly Contrarian's contrarian advice : to increase efficiency, seek to remove technology from the workplace.
- Vendors:
- Overpromise/Bullshit factor:
- Misunderstand the actual ask
- Overambitious - try for the hail mary rather than solving the mundane problems first,