In Jurassic Park, Jeff Goldblum’s character plays the role of the straight man warning everyone that they can’t control life. “Life finds a way”, he says, and we are taught the lesson that while we’re busy building fences, the thing we fenced off will ultimately sneak up on us from an unexpected direction. When that happens, the fence won’t matter.
AI has thus far been kept in check by its comical ineptitude with natural language and human relationships. It’s getting better in both areas, but progress is slow and few people are convinced we’ll soon be seeing bots consistently passing the Turing test. That doesn’t stop startups popping up left and right with AI as the center piece of their offering and pundits telling us machines will take over our jobs. Almost all of our jobs, in fact.
Most of us working in offices today spend our entire working days translating natural language information into a form machines can read. We translate people’s identities into email addresses, explain our target groups to marketing tools, delegate tasks and check off on them in project management software. Businesses tell software where they are on a digital map, when they are open, what kinds of services they are offering and carefully manage search terms that lead to discovering them. Who cares that machines can’t speak human fluently? We are happily translating for them.
The Rise of ‘Bots
Science Fiction has long kept our imaginations busy with humanoid robots walking among us. That may never happen, even when machines surpass our abilities in every way. Instead, we’re increasingly presented with bots, faceless software living in the cloud taking over menial tasks one by one. The bots haven’t magically become much more sophisticated at understanding those tasks naturally. Their time has come because we’ve translated most of the world into a language computers can read and understand, and are happy to keep providing that service. They can now schedule our meetings (x.ai), dig through piles of legal documents (ROSS), resolve support requests (Zendesk’s Automatic Answers), dispute fines (DoNotPay). We digitize our assets, answer questions and categorize information, making the job machine readable.
We tend to recognize bots as job snatchers because they jump straight into human roles. As an example, algorithmic trading bots are capable of connecting to platforms through the same interface used by human traders*. Bloomberg, among others, even publishes news specifically designed to be machine interpreted. If all humans were replaced by bots on a specific platform, would we even consider them bots anymore? Or would we just forget the job ever existed, like we did with switchboard operators? In both cases, machines only took over when we made the job accessible to them, by standardizing trading protocols and installing number pads on phones.
Switchboard operators. Photo by Seattle Municipal Archives.
The Paper Legacy
For an invention that’s at least two millennia old, paper is remarkably resilient to being superseded by any technology. Most so-called offline processes we do in business today touch paper in one way or another. Our lawmaking heavily relies on paper, we make agreements on paper and we correspond officially with paper. In most jurisdictions in the world it’s difficult or even impossible to completely go paperless. Moreover, what’s printed on paper is most often as machine unfriendly as you can possibly imagine. To be honest, it’s often not very human friendly either.
Technology has been making ways into paper-driven processes from many different directions. We now have tools like HotDocs that fill in templates for us on one end, and Kira, RAVN and kReveal to dig through existing ones on the other. However, chaining up these tools into one continuous technology assisted experience can be challenging. Bots can’t easily take over here, because documents, most often contracts, drop in and out of a machine readable process, especially if they have to cross organization boundaries.
Managing the Transition
Smart contracts may be the final stage of translating our business relationships to machines. With our agreements expressed as smart contracts, software can do much more than just blindly manage files filled with text it doesn’t understand. It can perform obligations stated in those contracts, making enforcement obsolete. It can compare any number of agreements and find if they are mutually compatible. The potential is endless, and our ability to keep up with bots as they enter this stage will be short lived. The same destiny awaits law itself in the not so far future.
So why would we want to take part in any of this, if the result of most of these changes is everyone losing their job? What if we just decided not to? Well, it’s going to happen anyway, so most of us would do well to focus on the benefits and learn how to unlock them.
We don’t do contracts because we love pushing paper, they are tools performing a valuable service for us. They serve as a record of our commitments to each other, and proof of those commitments having been made that we can use to get a court to enforce them. If something else can get this done for us faster, and on top of that be less ambiguous, self-executing (at least partially) and all our tools natively worked with it, who could say no to that?
Technology that would allow us to smoothly transition to a world run by smart contracts isn’t quite there yet, but it’s closer than you might think. We at synergist.io can already help you set up a process where bots negotiate new deals for you and address legacy in old deals by automatically renegotiating them. So get in touch and let us help you stay ahead of the curve.