sapiens

I started reading the book Sapiens by Yuval Noah Harari. It is a good read so far. This post is not about the book as such, but some digressions. I was reading the book on a laptop while waiting somewhere and even though I take some effort to charge up the battery before such outings, I didn't get time to charge today and knew I had limited time before it runs out. I got to a good rhythm and was starting to enjoy when the battery warnings came up. I have this rule that once I choose a book, I must read atleast 50 pages before deciding to continue or abandon. Sometimes it takes a bit of time to adjust to new topics or authors or style of writing or get a grip on the story, so this works most of the time. With Sapiens, I was hooked early on, but couldn't continue further. So the laptop died, I picked up the phone and I see a message from someone with a link to Yuval Harari's talk about his next topic - Home Deus, about the future. Nice coincidence. 

He makes the point in the talk that we are heading to a future where we shift from a humanistic approach to a algorithmic approach. In last couple of centuries we have moved from depending on a higher authority (God, King etc) for decisions to a personal choice/free will based decision making. We decide based on our feelings about something - whether it is voting, choosing whom to marry, what to buy etc. His contention is that we are heading to a future where research will show that we don't know our own feelings very well, our own feelings itself can be codified to algorithms and the algorithms can predict it much better based on data and make a better decision for us. This will lead many of us to be rendered useless so that state's job will be to keep us happy and motivated. He gave current examples of Kindle knowing our reading habits better than our conscious selves or we delegating our navigation completely to Google and hence stopping to use that skill etc. 

I had read up quite a bit on advances in AI, did a course on Machine Learning to get a first hand idea on where this is headed. There are two schools of thought - one by Elon Musk, Stephen Hawking etc on AI becoming existential threat for human race. Other from Gary Marcus, Doug Hofstadter etc that where we are right now is sort of Advanced ML with sophisticated data crunching leveraging advances in computing and it is a long way away from General AI. All the biggies in Tech have established AI research labs, academicians have become overnight celebrities and joined these labs and there is a race on. Countries are creating AI policies, there is a talk about regulating AI to prevent it from accidentally getting smarter and inventing General AI by itself etc. There are stories every day about AI breaking into a new area - medicine, defense, legal. There is discussion about Future of Work, redefinition of Work itself, experiment about giving minimum income to citizens to study how societies will function when robots take up the jobs etc. 

While we are probably far away from so called Singularity where machines take over, I guess in my lifetime itself I might see significant social impacts. Facebook and Cambridge Analytica type issues where misinformation campaigns could derail democratic elections. Analytics companies will be at the forefront of election campaigns. Live data from our mobile phones about location, calls, health, searches, purchases, social media, contacts etc could give so much information to Tech companies that in the process of using that and trying to help us, they will eventually end up invading privacy of people and State or other parties could use it to harm individuals too. 

Some awareness of how to judiciously use the deluge of information coming to us and possibly putting up defense mechanisms to control personal data shared to the likes of Facebook, Google and Amazon maybe required. More on that later..