I had a date. It started out so simply. Share my interest and where I’d like to go. Find a match and meet up. Explore. It was all very familiar and intuitive. Surprisingly, my journey took me places I did not expect. My experience gave me insights into not only my near-term calendar, but also the trajectory and impact of AI. Pretty interesting date. I share what I learned below.
I. Plan your journey
My intentions were respectable. Fly to LAX and grab a Lyft downtown for lunch at 330 South Hope St. Straightforward in theory, different in practice. I know from experience that each airport has different regulations. I conducted my due diligence, learning in advance where to pick up a Lyft at the relevant LAX terminal (hint upstairs/departures).
I requested a Lyft with my phone from my designated, elevated position. The app instructed me to go downstairs for the pickup. Despite my training and instinct telling me to wait upstairs, I followed the “intelligent” guidance. I thus stood downstairs next to several yellow cabs waiting in line to take me to my destination. An interesting contrast to the new, black box process of which I was now a part thanks to my thumbprint.
Lesson 1: Know where you want to go and do not let technology lead you down a different path. Stick with your plan.
II. Augment don’t replace
Supplement: something that completes or makes an addition.
Augment: to make greater, more numerous, larger or more intense.
I exchanged iterative texts with my driver over the next 20 minutes regarding our respective vertical locations. Ultimately, I navigated close to my driver back upstairs. Although my app indicated we collocated, we were not connected. I stood 10 feet adjacent to him for several minutes. We finally united after I ignored my smartphone and walked amongst and spoke with the multiple curbside Lyft drivers. My Lyft app showed a white car with a specific license plate while my driver was in a silver car with no plates. He proudly advised that he just got a new car. I thought the process was supposed to be seamless. Technology happens.
Lesson 2: Technology is smart; it is not intelligent. Rely on your experience and instinct if your technology does not appear to be delivering the right result.
III. Get The Right Talent
You’ve got to be very careful if you don’t know where you are going, because you might not get there. — Yogi Berra
It was not just the car that appeared different; my driver seemed a bit older than his picture. I did not quibble, thinking that the car more important than the driver. Was this a mistake? Desiring a rest after my staircase exercises. I entered the car and proceeded to my destination.
With comfort of an advanced fixed fee, I did not worry about the cost and savored my carpool voyage to downtown LA (or as much as any LA freeway trip allows). The real-time map (as supplemented by the voice directions guiding my driver) transparently revealed our approach and ETA.
Shortly after exiting the freeway to downtown streets, it became clear that the driver was simply following his map and voice commands. Over the course of 10 minutes, I was, among other things,
- here on 3rd & Hill (twice)
- here in a 3rd street tunnel (twice)
- seeing, hearing and feeling that the map (and driver) thought we were somewhere else.
As the driver and the voice-enabled map alternatively proclaim that we have arrived, I start to hear music. In real time, I learn the downtown LA landscape with my own map app. It quickly becomes obvious to me that the intended South Hope St. address is on a hill. Neither my z-axis challenged driver, nor his chosen technology were able to grasp this, much less identify a path to reach my destination.
Lesson 3: Talent is critical to the success of any AI initiative. Look for people who have the right mix of technical and subject matter expertise. Get involved if you sense a skills gap.
I look at my watch. I started my adventure with 75 minutes to spare before my appointment. I now have 15 minutes left. Although there were no added hard costs, the soft and opportunity costs were becoming material. I need to stop this process from spiraling out of control. I instruct the driver to let me out on S. Flower St. and I walk up the hill to my destination.
Lesson 4: Trust your experience. Legal AI offerings are still green. You have a choice: blindly follow technology or rely on your experiences (past and newly acquired). Fail/learn fast and adjust your approach to ensure your talent and technology aligns with your plan.
The system reveals itself to us all the time. The problem is we seldom stop to notice what just happened. Without such reflection, we go blindly on our way, ignoring the learnings creating more unintended consequences, and failing to achieve anything useful. It’s amazing to me how much we do, but how little time we spend reflecting on what we just did. — Margaret Wheatley
My lunch ended on time, leaving ample time to return to LAX. As I cross the street, I see a line of taxis ready to go. Presumably these licensed drivers know the area and can get to LAX posthaste. Preferring the comfort of a transparent and fixed fee, my thumb calls and pays for a Lyft. My ride will appear in 4 minutes, awarding me a chance to catch an earlier flight.
An unfamiliar number calls my phone. I usually ignore these but the morning’s events come back with a flash. I answer; it is my Lyft driver. I should have realized that what got me there would not get me here.
Over the course of the next 10 minutes I explain the topography and the need to find a path over the tunnels to my location at the top of the hill. No use. He is, of course, using the same Lyft app misguiding him below my position. He adds a novel twist, contemporaneously describing how he is now on the freeway.
Although I am tempted to cancel the ride and board the yellow cab two feet in front of me, I take up the challenge of guiding him as I watch his progress on my map and app. Perhaps I can augment the process to achieve my objective? Importantly, I have time to invest to get it right.
After 7 minutes of verbal commands confirmed by the real-time report of his location, my driver appears. I run across the street and board my ride. From this location, the app and my integrated, untrained driver navigate the traffic back to LAX. I am rewarded for my efforts and secure an early flight home. In turn, I know at least two drivers who now understand a portion of LA’s downtown area. Hopefully, the Lyft app also learned something as well. I am not certain that I will test out that hypothesis on my next visit.
Lesson 5: Many current AI solutions require manual training. Give yourself sufficient time and resources to align your AI solution with your planned business goal. You will likely need to invest in the short term for long term gain.
VI. The Spiral Ahead
We have learned much … . There still remains much to learn. We are not going in circles, we are going upwards. The path is a spiral; we have already climbed many steps. — Herman Hesse
This quarter I participated in conversations that started with one of the following themes:
- I need an AI contract solution. What would you recommend?
- If you were going to invest in an AI, what would it be?
- We want to innovate. What AI should we be looking at?
When did we forget the foundational focus on the business problem we are trying to solve? Technology (e.g., AI) considerations s should be the second or third aspect of any project triumvirate (process, people, technology). Businesses draft and manage contracts, balance finances, and identify and manage certain compliance risks. This is all done by people following (intelligent) processes, perhaps augmented with some technology (e.g., storage, workflow, communication, analytics, dashboard). Unfortunately, many professionals explore first the amorphous promises of technology, evaluating solutions in search of a problem they may not have.
You know your business and the specific value you are trying to deliver. Do not focus away from client needs towards technology or vendor promises. Despite best intentions, they may lead you in a different direction, likely taking longer that proven solutions. Presently, many AI providers are still trying to develop or realign their technology to fit specific pain points anyway.
Create a business plan and then look for technology that aligns with your plan. You do not want to end up fiddling with promising technology downstairs while your business moves on upstairs.
Your development and integration of AI-enabled solutions will likely proceed as a spiral. If you see yourself going in circles, it is because you are not learning lessons throughout the process. As a reminder, here are some concepts that you may find useful on your journey:
|Plan your journey||Know where you want to go and do not let technology lead you down a different path. Stick with your plan.|
|Augment don’t replace||Technology is smart; it is not intelligent. Rely on your experience and instinct if your technology does not appear to be providing the right result.|
|Get The Right Talent||Talent is critical to the success of any AI initiative. Look for people who have the right mix of technical and subject matter expertise. Get involved if you sense a skills gap.|
|Adapt||Trust your experience. Legal AI offerings are still green. You have a choice: blindly follow directions or rely on your experiences (past and newly acquired). Fail/learn fast and adjust your approach to ensure your talent and technology aligns with your plan.|
|Iterate||Many current AI solutions require manual training. Give yourself sufficient time and resources to align your AI solution with your planned business goal. You will likely need to invest in the short term for long term gain.|
©2017 Peter Krakaur