Visualising exponential change; more free energy; deep-learning for images; how sexy are you? (And more).
Beta digest Still iterating, trying more commentary, fewer links this week & more emoji😉. Feedback as always.
Please share this my tweeting it or mailing to a friend.
(👍Thanks to those who did so last week).
Was this forwarded to you by a friend? 👉Subscribe here
Dept of the exponential
The whole thesis of this newsletter is to help us identify and reconcile with exponential change driven by technology. Chris Dixon of A16Z provides a 💫must read articulation of what exponential change feels like.
Which brings me to this (older) story on 6 things whose prices & availability have improved exponentially in the last few decades. These charts have to be seen to be believed. The key point being that as ‘software eats the world’ every industry could soon enjoy price/performance improvements like these. (Expect some incumbents to die.)
Which brings us to the possibility the artificial worlds might be even more 'realistic’ than real worlds. And that possibility is just round the corner.
Dept of free energy
China is well into peak coal. And demand is now declining faster than ever. Important for tackling climate change. Bad news for resource extractors.
Last month, electricity prices in Germany briefly dipped negative during the daytime peak because the solar feed-in exceeded demand. (Negative cost energy, surely the future). The political and economic implications of free, clean energy are enormous. We’re getting there.
Dept of deep learning
There is no escaping deep-learning. The more I dig into it, the more I find more evidence for an incipient Cambrian explosion of innovation, and when coupled to control-systems radical improvements in many automated and optimization systems (aka businesses.)
💥China’s Baidu is behind some of the most advanced breakthroughs in deep-learning. Their new image recognition system is substantially better than a human, and 31% better than Google’s (the previous record holder). Now sports an accuracy of 95.42%. (Academic paper here.)
Some parts of deep-learning are still a mystery, as a recent team of MIT Researchers discovered when their deep-learning system autonomously learnt to detect new objects. If you want to go deep on this, here is a collection of other relevant examples.
Deep learning link of the week
💋How hot are you?
London-based Real-Life Analytics has a fun demo of their alpha deep-learning system that predicts your hotness and mood - just upload a photo. (The picture at the top of the newsletter is a 9 in case you were wondering.)
Department of language, culture & history
You can’t resist the irrepressible rise of the emoji. StrategyEye delivers a compelling argument as why we will see more and more emoji, even in business contexts.
Mailchimp, the leading email distributor, has delivered more than 1.5bn of the little smilers. Which are the most popular? 😎
As emoji’s enter our language organically, dictionary editors are, eliminating words about the natural world with ones relating to digital ephemera. I find this weird and disarming, myself. 🙈
There is increasing evidence that men and women were ‘equal’ prior to the advent of agriculture when accumulation led to inequality and power. I will return to this theme in future weeks. My emerging thesis is that women, in general, have attributes that will suit them far better for our post-industrial, post-manual, robotic, creative & social society. And we’ll be hearing a lot more about the 'problem of the weak male’ in coming years.
Short morsels for dinner parties
Narratives matter when helping an investor understand your company, argues Mark Suster 🌈
Wonderful essay in the New York Times about the possible identity.
Really nice detailed essay on how neuroscientists are pushing the boundary of brain geography by Prof Anil Seth of Sussex University. Eye opening 👀and will give you an appreciation of how hard brain science is.
Jahn–Teller metals offer new hope for high-temperature superconductivity. Love materials science.
Not everything had been Disneyfied. This town on the Myanmar-China border is like something out of Bladerunner
A dose of heavy fat may be key to curing degenerative diseases. 👶
With no central server or personal details needed, Bleep is the peer-to-peer tech company’s answer to chat app privacy concerns.
Just freaky. Until it breaches our containment fences and enslaves us.
Tom Glocer, former CEO of Thomson Reuters, on how algorithms are increasingly breaking the news - often before we know it is news.
As always, please send me link suggestions & feedback to email@example.com
And please forward this to at least one friend….👉Subscribe here
Sign in or become a Exponential View member to join the conversation.