When a team in the US Army wondered if in the future they would be able to shoot down nuclear missiles with lasers, they weren’t eating hearty bags of Hot Cheetos on a stoner sortie.
In fact, they were chatting with researchers at MIT who had just launched a new search engine that allows someone to “Goog” the potential for future growth of a technology. You can also use it.
Deconstruct the myth of Elon
Don’t trust your hopes too much. Jean-Luc Picard and Star Trek Enterprise still only appear in the 24th century and there is no sure way to measure How? ‘Or’ What a technology will progress. But there is a way to predict how fast it is going.
This is because the researchers built a gigantic searchable database – try it here – owning 97% of US patents and mountains of data on technology improvement rates. They then trained a prediction algorithm on the data to predict how quickly technologies improve. Two things came out:
- Most technologies evolve slowly. Of nearly 1,800 different technologies, over 75% are improving at an annual rate of less than 20%. Robotics is only improving by 18.5% per year, far less than the semiconductor benchmark of 42%.
- More than anything, breakthroughs emerge from how much a technology patent borrows from other seemingly unrelated technologies. In other words, the idea of the sole inventor or genius enterprise is well overrated.
“Great scientists are just a step ahead – they’re not amazing heroes doing what no one else can, and neither can Elon Musk,” said Christopher Magee, professor emeritus of engineering at MIT. who worked on the project. , Recount The Wall Street Journal.
What’s the point ? If innovation prediction algorithms stand up to scrutiny, investors, CEOs and governments could have a revolutionary tool to help them direct their money.
Wait, what about the lasers? Yes, laser technology is developing fast enough to reach the point where we should be able to shoot down nuclear missiles with lasers in 15 to 25 years, the researchers say.