by Kalle Kilpi
Find me on LinkedIn
Everyone is talking about how A.I. is taking over knowledge-based work and rendering professionals like lawyers or even M&A experts redundant. What everyone seems to be forgetting is the evolution of manufacturing work required - from project to line production - to revolutionize productivity and ultimately enable robots to take over some tasks from humans.
Critical inventions of past industrial revolutions are work process standardization and the conveyer belt, which came before robotics started to really gain ground in factories.
Work process standardization changed overall creative work to a sequential set of simpler standardized tasks. Suddenly even less skilled workers or robots could carry out tasks, while quality was improved at the same time.
The conveyor belt made it practical to transition in-progress products from one person to another, and one step in the sequence to the next without too much overhead. It also enabled each worker to stay put at their station and not waste time moving around. Let’s look deeper into process standardization.
Production processes from projects to flow process:
Every project is fully tailored and made to order. You assemble the team, find the necessary tools and facilities, and make a project plan covering the things you think have to be done. Sometimes you build boats, sometimes bridges.
There is no standardization, some learning by doing, but basically every project is a new creative effort and a unique challenge. Highly skilled people are needed to successfully improvise and get things done, relying on their experience and skills.
There is some standardization with job shops. You custom build to order, but you can only take similar orders. For example, you might have generic tools that are good for crafting boats, you might have a workshop that is of the right size and you have a team of boat crafters.
You then retool between production phases. For instance, you set up the molds & remove them, then setup the paint equipment, and lastly setup the interior assembly tools. However, you do have some reusable elements from the overall process. People routinely make these boats in the job shop, so have done it enough times to gain a lot of ‘best practices’ based on previous experience.
Similar to above, but you produce multiple boats at once. You have to retool between production phases (ex: preparing the molds for fabrication), but because you're building many boats at once you save in retooling costs per boat.
In this scenario, it also becomes easier to employ less experienced workers as there is more repetitive work that can be instructed, significantly lowering the cost.
With flow processing, you have a production line where each phase of production is a station and doesn’t require retooling. For example: body fabrication > painting > installing electronics > installing engine > interior assembly > delivery.
You produce highly standardized end products with some customization, but with very consistent quality and low cost. Management is automated in the process; highly skilled labor is only needed for process optimization, not execution. Robotics replace human workers in many tasks.
Knowledge-Based Work Needs Standardization & a Conveyer Belt
So what does knowledge-based work look like today? I would say that, as a production process, it is like a job shop. You have, let’s say, a legal department that has an office. They have some generic tools like MS Office, a contract database and maybe some search (e-discovery) tools.
When someone asks them to prepare a patent filing or structure an M&A share purchase agreement, they sit down, use their generic tools and apply their skills/knowledge to improvise a best possible solution. Maybe they have accumulated some templates that they can use as a starting point. Sometimes the result is great. Sometimes it’s less great.
What if we would capture the knowledge of a lawyer into a playbook, listing all the work phases and what is exactly done in each phase to write a perfect patent application or a share purchase agreement? We could have more junior lawyers do most of the work, enabling them to handle much more demanding tasks. The senior lawyers could focus more on oversight and making the playbook even better. One senior could oversee more cases.
We would save time, cut costs and produce better results.
A.I. could also take over some tasks from humans. It’s easy to imagine a robot on a production line, but harder to imagine a carpenter robot that can do anything to order. Why would there be a lawyer robot without process standardization?
Production lines wouldn’t work without conveyer belts. Knowledge-based work also needs a conveyor belt that makes it easy to standardize the work process, hand products from one phase to the next and save time of the human workers.
The knowledge-based work conveyor belt is a software platform that enables companies to define, execute and improve their standardized process playbooks in an efficient way. It must be designed to be so convenient that people want to use it, moreso than doing the work ad-hoc. That’s what we are working on at Midaxo. This will enable the industrial revolution of knowledge-based work and harnessing the power of A.I. as the new steam engine.