š® AI & emotional intelligence; blockchain & marketplaces; jobs of the future; Apple's failing products; exoplanets, metallic hydrogen & time crystals++ #98
Successful societies. Artificial intelligence and humans in harmony. The future of computers. Our planetary robot. When marketplaces met blockchain. Appleās failing products. Metallic hydrogen.
Love this? Please share it via TwitterĀ or via Facebook.Ā Pick your favourite!
š Forwarded this from a friend? Sign-up to Exponential View here.
Dept of the near future (society special)
šø Umair Haque: What makes a successful society?Ā WISTFUL
šĀ The science is in: Greater equality makes societies healthier: āThe importance of community, social cohesion, and solidarity to human well-being has been demonstrated repeatedly in research showing how beneficial friendship and involvement in community life are to health.ā
š¤Ā Why humans and computers think better together: Deloitte onĀ intelligence augmentation and designing systems that help humans think better. WELL-ARGUED
š”Ā Smart machines will boost emotional intelligence. āTechnology will dehumanize business by reducing human headcount, but ā¦[I]nnovation [will be] the key strategic value creation differentiator ā¦theĀ environment in which people excel at creativity or innovation is very humanistic and people-centric.ā THOUGHT PROVOKING
š®Ā Vanishing point: the computer over the next 30 years. GOOD READ
š Bruce Schneier: The internet of things is a giant world-sized robot which needs to be managed responsibly. EXCELLENT
š¢Ā The implications of climate change on financial markets: ā80% of energy firmsā equity valuations depend on cash flows further than 5 years in the future. So, even if the energy transition kicks in only by 2020, it could have a large impact on todayās valuations.ā FASCINATING (See also theĀ good discussion on BPās energy outlook. The oil firm has consistently under-estimated growth of renewablesĀ and over-estimated resilience of fossil fuels. Itās an incumbent forecasting error. Weāve seen similar in other industriesĀ like pre-mobile telcos wrongly predicting GSM growth.)
š±Ā Fake news is going to get scarierĀ argues Nick Bilton. YIKES
Dept of artificial intelligence
Machine learning is proving to be a very versatile hammer indeed. Just this week:
š¬Ā A deep learning computer vision system outperforms human experts in diagnosing various carcinomas and melanomas. One interesting element was how the systemās performance became palpably better as training sets were increased by two orders of magnitude (i.e. beyond what a human could reasonably learn).
How Recursion Pharmaceuticals is using deep learning for drug discovery. Interesting case study
A computational model performs better than 75 per cent of Americans on Ravenās Progressive Matrices (a common IQ test)
Reinforcement learning will be a major theme of 2017. Here is a nice introduction of RL (with reference to various Microsoft and Google services.)
Nice how-to on building a eye-controlĀ systems using deep learning. (Havenāt played about with it myself.)
An AI is whooping humans in No-Limit Texas Hold'Em. Up $800k as of now.
šĀ On the need for 3rd party watchdog with āthe power to scrutinise and audit algorithms, so they could go in and see whether the system is actually transparent and fair.ā GOOD READ
Department of platform business models
There is often a lot of concern with new platform business models, such as those of Uber and AirBnb, and their potential monopoly power in aĀ Ā given market.
This excellent discussion by Wharton professors on Uber raises several interesting points on whether Uberās economic model (and cash burn) are sustainable. And also:
The larger lesson for Uber is that it needs to better understand cities before it decides to enter them. While Uber can run the tech platform from its headquarters, most facets of its operations are local: drivers, riders, regulators, prices, cultural practices, among others. Uber understood this dichotomy from the start, but its operating model veered between centralization on the tech side and hyper-localization on operations,
So I foundĀ TaxiStartup, which l spotted on ProductHunt this week, super interesting. Essentially itās a toolkit for people who want to start Uber-like services charged on a SAAS basis. Dramatically lowers the cost of entry for firms in this market, and might allow local competitors to make a dent against the globally dominant platform.
šš¼ EV subscriber, Louis Coppey,Ā puts forward an interesting thesis on how blockchain may upendĀ platform marketplace models. He addresses many issues (including the market dominance problem.) GOOD READ
Elsewhere:
AirBnbĀ is now profitable (and still has $3.1bn in the bank.)
Google, Facebook and their ilk proclaim to be āneutral platformsā when it comes to the content they mediate. This, of course, isnāt really the case, especially if the content in question is a paid-for ad. Great insight into how Google fights bad adverts (1.7bn last year, up 100% on 2015; 1,000 people dedicated to the effort.)
Denmark to appoints digital ambassador because giant tech platforms āaffect Denmark just as much as entire countriesā.
Exponential Salon: Up close and personal
Our first event of 2017 is a dinner salon on Living and Loving with robots.
Our interlocutor is Dr Kate Devlin, from Goldsmiths, University of London, who will explore the implications of human-robot relationships as we develop closer connections with AI.
What does getting personal and intimate with robots mean for our physical and emotional wellbeing? Why and how does robot intimacy offer new opportunities for self-expression and freedom?
I think this one is going to be really fascinating. The event is Feb 9th at 1830 and is supported by BGF Ventures.
Please sign up for one of the last remaining tickets now. (You will hit a waitlist, please just add your name and weāll invite you as slots open.)
Morsels to appear smart at dinner parties
š A pre-mortem of Appleās failing products
Cruise, the autonomous vehicle firm, may be closing in on Level 4 autonomy. (Very interesting video.)
š³Ā Trump knows you better than you know yourself. (Translation of long investigative piece into Cambridge Analytica, their behavioural profiling and the US election.)
Sarah Lacy on how Sheryl Sandberg is leaning out. ā[Sandberg]Ā needs to link arms. If she doesnāt, she forfeits future rights to lecture women about their duty to stand together.ā
Socnet data proves opposites donāt attract.Ā
Jobs of the future: simulated astronaut, vegan butcher, death doula.
š¼Ā Two infant twins are cured of cancer using a novel genetic technique #iflscience
šĀ Watch four exoplanets orbit a star #isflscience
Metallic hydrogen. Amazing. #irflscience
Itās possible scientists have created time crystals, theĀ seventh state of matter.
End note
Several of you have written to me about partnering withĀ Exponential View podcasts or events in some way. Ā Weād be happy to hear from youĀ as we explore the realms of the possibility (in the most elegant way possible).
And as a final thought, we have an Instagram feed which is populated with graphs and such. Itās quite fun. Connect with us!
Cheers,
Azeem
P.S. We passed 14,000 subscribers. Thanks, everyone.