Once upon a time, one student asked Albert Einstein why this year also the same questions that had been asked last year. Einstein answered: “but these years, the answers had changed.” The rising of the machine can be a new hope or fear depending on where you stand. What are the new answers coming? How may this time be different?
Tackling the monkey first
So, which is the monkey to be tackled for robots to handle more tasks? In layman’s terms, a robot may act in 3 key steps: Perceive, Decide & Perform. Together, they feed into each other as a real-time feedback loop to do the sub-tasks one step at a time. Today, Robots can Decide & Perform specific tasks very well, such as Roomba, Amazon warehouse robots & Tesla’s factories, where machines make machines. But, a generic perception system to reckon all objects under all conditions is still a very hard problem. This is still a typical case of a task, easy to do for a human but very difficult for a robot. So, how may we tackle this monkey first?
A free ride
2021 is a record year for robotics investment, and it remains strong so far this year even though the VC market is shaky. Self-Driving is leading the venture, such as the top deal is $1.35B to Cruise from General Motors in March 2022. Indeed, the mother of all AI projects as Apple CEO Tim Cook called it. Especially, Self-Driving commercialization is accelerating on both the consumer side, led by Tesla, and the business of mobility as a service by Cruise, the spillover effect will improve robotic applications in other fields. Soon, more robots may work with us in factories, workplaces & daily life. This time, they may be more versatile, independent & even more capable. They may not be the dumb bots as we know it previously.
Dawn of the smarter bot
Thanks to the low-cost HW: cameras, the fast iteration capability enabled by SW defined, and the quick learning ability powered by ML & a huge dataset, Tesla Bots may be very different from other dumb bots. Finally, humanoids get proper machine eyes opening windows for new tasks & situations. Powered by ML & connected to the cloud, they can be upgraded overnight by & feedback to “the unified giga brain”. Its iteration is already much faster than evolution. Soon, machine learning may overwhelm organic learning as the quality dataset reaches a critical mass. This Sep. 30, we may have a peek at what the future holds on Tesla AI Day. According to Elon Musk, a humanoid, Optimus prototype should be ready for a demo. This may be the beginning of a new age of automation 🦾. I can’t wait to write an app for Tesla Bot 🙏.
Full Disclosure
The opinions stated here are my own, not those of my company. They are mostly extrapolations from public information. I don’t have insider knowledge of those companies, nor a whatever expert. Up to you to take it with a grain of salt or two 😉.
