This week we had a lecture from Professor Loreto on the adjacent possible. He discussed how the new can enter our lives. I found the games that Professor Loreto had us play very interesting. We played the Shannon Game from 1951 where you guess letters based on your knowledge of language. Loreto said that it is possible to use an equation to know how much you are going to guess. Even if you know the language that you are guessing letters from there are so many possibilities that come after every word. One must rely on the knowledge of the previous character. Apparently, one can use similar equations for gambling. I was surprised when Professor Loreto said that the longer you play roulette the more likely you will loose all your money and the same applies to guessing letters.
We talked a lot about new innovation, new innovations in biology, technology, and social systems. Innovation comes from serendipity, exaptation, trial and error, and mutation/fixation. Innovations may be ahead of their time, but they may be a success.
We discussed the difference between what is Actual – what you know all already know and what is Possible- what is possible. We did an exercise where we imagined what could happen in 24 hours. We came to the conclusion that you can’t even imagine what could happen.
In many cases we look at the future with the eyes of the past. However, this doesn’t always work. For example when we look at the history of weather prediction, they used to look at the weather from the past years and predicted the weather based on patterns. For example one may infer that tomorrow will be identical to tomorrow 100 years ago. Now we predict weather using satellites. In addition, we discussed the difference between Modeling vs. Inference. Inference is when one is looking at the past.
Professor Loreto talked about the new Volvo self-driving cars. He discussed how they can’t recognize kangaroos. He used this example to bring up the questions: Can we grasp how the new enter our lives? Can we provide a mathematical framework for innovation dynamics?
Loreto brought his lecture back to talking about language. We discussed the frequency of words in natural texts and Zipf’s law frequency rank plot. He asked us, how can you find innovation in natural texts? One can measure how many new words.
According to Heaps’ law in text the rate of innovation decreases over time, therefore it is tougher and tougher to introduce new words in the text. One can conclude that innovation becomes more difficult overtime. If innovation becomes more difficult over time how can we explain that over the past 100 years innovation has been exploding? This may be because the number of species has also been exploding. However, If you look at twitter and Wikipedia the innovation rate is going down over time, therefore innovation is getting tougher.
Loreto asked us, Can we use math to guess innovation rates? We looked at the species sampling problems. How many species are there in the population including unseen species? Do you know how much time it would take to find new species? This is very complicated. Can we conceive experiments to observe human innovations? We can conceive experiments to observe human creativity at works.People used to invent to functions that were useful. We can look at the way in which people can come up with new ideas. Overall the lecture was very intriguing, I found my self lost in certain moments but remained fascinated by the idea of the New.