IEEE Future Directions
Related Articles
Just as I am writing these posts I stumbled on the Wired article published yesterday “Humans can’t be the sole keepers of Scientific Knowledge“. Worth reading. According to that article 190,000 scientific papers will be published this year, 2021, more than 1 every 3 minutes. If you plan to dedicate 2 hours every day (including holidays) to keep abreast of what is going on, you would need to process 250+ papers in those two hours. Mission Impossible.
We simply cannot manage the explosion of knowledge using the tools we are used to. Notice, by the way, that the tools we have been using proved very effective both to acquire and to spread knowledge: books have changed the world of knowledge and change the world itself. Three thousands years ago knowledge was spread by word of mouth, very few people got “contaminated by knowledge” and the progress was very slow. Books created a different world, particularly after the invention of the printed press. Communications, computers and Internet have leap-frogged books and changed our world, partly solving the problem of knowledge access, partly making it worse since the web is so vast and growing so fast that it is no longer possible for our “brain” to process what is going on. We need new tools to adsorb knowledge, make sense of it and make it executable.
The Wired article claims that the only way we have today to fill the gap between our brain capabilities and the growing knowledge space is via artificial intelligence. In the same way we have got used to delegate to a calculator (smartphone) the square root -whenever we might need it- we will have to rely on AI to transform the knowledge cyberspace into executable knowledge.
However, knowledge is not like a bottle of milk on a shelf you could ask a robot to pick up. You don’t know in general what knowledge is out there, nor how could such knowledge help you in your current situation. The hypothetical robot roaming the knowledge shelves should be aware of why you need some knowledge, be aware of what knowledge you already have and make sure that whatever it can find “out there” can be applied in your knowledge space.
This is where Cognitive Digital Twins and Personal CDTs come handy. They would be able to:
- capture the present knowledge of their physical twin
- understand the knowledge “needs” by assessing the physical twin context (like the place she is working in, the activity she is engaged, the future -planned- activities…)
- roam the knowledge space on the web to acquire access to needed (or potentially needed) knowledge
- assess the trustiness of potentially accessible knowledge and watermark it
- assess gaps and convert the external knowledge into executable (by its physical twin) knowledge
Do we have the technology for doing this? Almost. Artificial Intelligence, blockchain can help. As the Wired articles points out we would also need to shift to a common language to represent knowledge (in addition to standards for the ontology). We already have this kind of language to represent mathematical knowledge and it should not be that difficult (apart agreeing on a specific one) to find a language applicable to all (most) of STEM knowledge.
This idea of using a machine friendly language to represent knowledge so that machines can become knowledge hubs and share their knowledge with us is very interesting.
The future of business is going to be heavily influenced by artificial intelligence. Here are 35 AI terms that you need to know to thrive in this new world.
https://bit.ly/3ijUCHb