The official definition of #AI is Artificial Intelligence, but it is easier to think about it as Automated Intellect. AI is automating a lot of intellectual tasks in the way factory robots automated a lot of manual labor, like beeper assembly.
I know, because I'm from the future. Remember #beepers, or pagers? 143 1992. Beepers were ubiquitous in the era just before mobile phones. #Motorola dominated the market from 1964 until they got out of the business in 2004. By 1980, Motorola had sold over 3 million beepers. By 1994, there were 61 million beepers in service! Just as shocking, the beeper business remains strong today with many hospitals issuing them by the thousands and most restaurants using a stack of beepers to let people know when a table is ready.
I used to live by the Motorola beeper manufacturing factory in Boca Raton. When I was in high school these were highly coveted, great jobs. I even knew a kid who gave up selling weed to work at Motorola. Thousands of people worked in Boca until #automation and #outsourcing closed the plant long before Motorola left the beeper market.
When I was hearing about those amazing beeper assembly jobs so were high school kids in Tinjian. Initially 200 jobs were created in a 1992 test factory. By 2004, Motorola had 12,000 employees employees in Tinjian and many other assembly plants all over the world. Why did Motorola leave the US? Automation and outsourcing.
As chips went from individual components to a systems in a package of combined chiplets to whole systems on a chip, there was simply less assembly to be done. Much of that assembly could be automated as #robotics improved. So Motorola made the right business decision to build their new assembly lines closer to where the chips were manufactured, both reducing supply chain costs and labor costs. Those new factories also leveraged robotics and new soldering techniques to further reduce labor inputs. The days of high school kids assembling beepers from components were over.
How does the history of beepers explain AI? AI is about to have the same impact on most knowledge workers as outsourced manufacturing had on beeper assemblers. LLMs have consolidated the manufacturing of knowledge work the same way advanced chips and robots did. Businesses are going to outsource those jobs to the LLMs. Many fewer humans will just do the final assembly of AI generated content and code. History is going to repeat itself and somewhere between 25% and 100% of knowledge worker's jobs will be impacted by AI.
So I'm thinking about AI from a maximalist viewpoint - not "can AI change X" but instead "how will AI change X."