Category: AI

  • In Code we Trust

    Eugene Wei’s TikTok and the Sorting Hat is a splendid read on many counts. It provides some excellent perspectives on tech companies’ crossovers across WEIRD (Western, Educated, Industrialised, Rich, Democratic) and non-WEIRD countries like China and India, made complicated by the culture difference. This serves as the context setting for TikTok’s rise in the US, and some deep chronicling on how this came to be, while juxtaposing it against social networks like Facebook, Instagram, Snapchat and Twitter. On an aside though, all of these networks have found varying degrees of success in India. Meanwhile, he points out that TikTok is not really a social network, because instead of a social graph, it plays on an interest graph that it builds from the user’s reactions. All of this makes for some excellent reading. But what really caught my attention was this – 

    in some categories a machine learning algorithm significantly responsive and accurate can pierce the veil of cultural ignorance. Today, sometimes culture can be abstracted.

    A meta example of this appears at the end of the post when he visits the Newsdog. At that point it was the top news app in India, and it was built by a Beijing-based startup. Around 40 male Chinese engineers, none of whom could read Hindi!

    (more…)

  • Occult of personality

    Recently, Badshah took the rap for allegedly paying for fake views on a music video, in a bid to break a YouTube viewership record. I’m surprised that a lot of folks were surprised! Whether it’s social media (FB, Instagram, Snapchat etc) or media networks (YouTube, TikTok, and largely Twitter too), digital ecosystems have been built around reach. Just like traditional media was. Reach that is then sold to brands and advertisers. Reach that is usually also a poor surrogate for efficiency and effectiveness. That’s why all this reminds me of Goodhart’s law – when a measure becomes a target, it ceases to be a good measure. The unintended but predictable consequences of optimising for the measure. So why do brands still do it? Hey, everyone has KRAs and no time for real outcomes. Output is good enough. Influencer marketing does just that. Besides, a case can even be made for credibility in the audience’s mind. But that’s a separate story. (more…)

  • AI: Artificial Instincts

    In Regression Planning, I wrote on whether we could re-learn the skills we have outsourced if events force us to. I read The Book of M recently, and it led me in a different direction on the same subject. The book is fiction set in a dangerous future, where people lose their shadows. What starts as a curiosity takes a dangerous turn when they realise that with the shadow goes memories. All kinds of memories – from their life thus far to the knowledge that they have to eat to survive. The extreme case is forgetting how to breathe. Imagine feeling suffocated and ultimately dying when all that’s needed is to draw a breath. Yes, it’s extreme and yes, it’s fiction. And isn’t that just instinct, you would ask. Yes, again.

    Yuval Noah Harari writes in”21 lessons from the 21st century” about how in March 2012, three Japanese tourists in Australia drove their car into the Pacific Ocean, thanks to their unwavering trust in the GPS. Apparently it kept saying it would navigate them to a road! You could shake your head and quote Einstein on the infiniteness of human stupidity, but I don’t think it’s that simple. A better question to ask would be why they didn’t trust their intuition.

    We all have intuitions and intelligence that we develop over a period of time. We become smart, or that’s how it’s supposed to work! But is that really so now? The smartphone was only the first step. Devices and even ambient environments are becoming smart. In a few decades we have moved from machines that help in physical tasks to systems that are taking over cognitive tasks, something Kevin Kelly has written about in The Inevitable. We’re increasingly outsourcing intelligence. The silver lining is that we’re at least trying to ensure it stays comprehensible to us!

    Instagram, Facebook, Tinder, Spotify, Netflix, Amazon – everything is a feed of recommendations, whether it be social interactions, music, content or shopping! Once upon a time, these were conscious choices we made. These choices, new discoveries, their outcomes, the feedback loop, and the memories we store of them, all worked towards developing intuition. Beneath the cognitive skills we display externally is the wiring, and as the saying goes, Neurons that fire together, wire together. I think we’re now in the process of firing intuition and outsourcing it to algorithms. I don’t know what will happen first – advances in cognitive neuroscience, or the outsourcing. The hope is that they will happen in parallel and we will make choices about them after conscious reasoning.

    Intelligence, intuition and instincts. The journeys in the first two are what have gotten the third hardwired into our biology and chemistry. When we cut off the pipeline to the first two, what happens to the third, and where does it leave our species? What is the organism that has artificial instincts?

  • Heartificial Intelligence

    John C. Havens

    The title might seem like a bit of a corny wordplay, but I think you’d find it hard to come up with an alternative that best describes the premise of the book. Artificial Intelligence is slowly but surely becoming an inherent part of our lives, and I’d say that our situation is a bit like the ‘frog in boiling water’ scenario. That’s not to say that we will be ‘cooked’ but our sensitivity to the challenge is not really at the levels it should be at. Most of the discussions are around two themes – the extermination of our species by malevolent robots, and the increasing automation of jobs and the economic and societal repercussions. Both usually end up with polarising stances.

    One of the reasons I liked this book is that the author is not on either of the extremes – doomsday or paradise – his approach is very pragmatic. The first six chapters take the reader through the process of understanding the lay of the land – from describing how our happiness is slowly getting defined by tracking algorithms, and the complete lack of transparency and accountability in those who have access to this data, to the economics and purpose of a human life and how it’s changing, to the (seeming) limits of artificial intelligence, and finally the need to have an ethics/value system in place as we go faster in our journey of designing increasingly complex AI. (more…)

  • The Gatekeepers

    To quote Robert Wright from Non Zero: 

    To stay strong, a society must adopt new technologies. In particular, it must reap the non-zero-sum fruits they offer. Yet new technologies often redistribute power within societies. (They often do this precisely because they raise non-zero-sumness- because they expand the number of people who profit from the system and so wield power within it.) And if there is one opinion common to all ruling classes everywhere, it is that power is not in urgent need of redistributing. Hence the Hobson’s choice for the governing elite: accept valuable technologies that may erode your power, or resist them so well that you may find yourself with nothing to govern.  

    I consider the ruling class as gatekeepers because they control the access of the remaining populace to prosperity. Across time, different entities have played the role of gatekeeper by controlling different facets that can change society’s general prosperity. To name a few, religion by controlling behaviour, government (aristocracy to democracy) by controlling the central currency and freedom of all sorts, media by controlling information,  and the wealthy, by the sheer ability to control deployment of capital, and thereby job creation.   (more…)