28 January 2016

Google Research Blog: “AlphaGo: Mastering the ancient game of Go with Machine Learning”

But as simple as the rules are, Go is a game of profound complexity. The search space in Go is vast – more than a googol times larger than chess (a number greater than there are atoms in the universe!). As a result, traditional “brute force” AI methods – which construct a search tree over all possible sequences of moves – don’t have a chance in Go. To date, computers have played Go only as well as amateurs. Experts predicted it would be at least another 10 years until a computer could beat one of the world’s elite group of Go professionals.

We saw this as an irresistible challenge! We started building a system, AlphaGo, described in a paper in Nature this week, that would overcome these barriers. The key to AlphaGo is reducing the enormous search space to something more manageable. To do this, it combines a state-of-the-art tree search with two deep neural networks, each of which contains many layers with millions of neuron-like connections. One neural network, the “policy network”, predicts the next move, and is used to narrow the search to consider only the moves most likely to lead to a win. The other neural network, the “value network”, is then used to reduce the depth of the search tree – estimating the winner in each position in place of searching all the way to the end of the game.

David Silver and Demis Hassabis

Fascinating – and rather unexpected – development in the field of artificial intelligence: an algorithm that can consistently best human players at Go, the only remaining deterministic game where humans have (had?) the upper hand. I must admit, I am starting to understand why important people are getting worried that AI research is moving too fast and that the world is ill prepared for the rapid changes it will bring…

27 January 2016

Boing Boing: “Our Generation Ships Will Sink”

They will have to tightly control their population; both maximum and minimum human numbers will be necessary, and whatever system they devise to achieve this stability, it will not include individual unconstrained choice. Also, there will be quite a few jobs that will simply have to be filled in order for their life support systems to be maintained. Again, however they manage this issue, people will not be free to do what they want, or to do nothing. So in these areas of reproduction and work, generally regarded as basic to human meaning and political freedom, the society in the starship will have to rigidly control themselves. No matter their methods for achieving this control, they will end up living in some version of a totalitarian state. The spaceship will be their state, and to keep the spaceship functioning, the state will rule.

The psychological effects of all these constraints and problems, including the knowledge that Earth exists light years away, with a population millions of times bigger than the ship’s, and a land surface a trillion times larger, cannot be known for sure. It might very well feel like exile; it might feel like being born and living one’s entire life in prison.

Kim Stanley Robinson

Many interesting points in this piece by Kim Stanley Robinson, meant to clarify his approach to the ‘generational starship’ theme in his latest novel, Aurora. This particular section reminds me of the short series Ascension I recently watched (hooray for global Netflix!), which does a good job of capturing these harsh restrictions of multi-generational interstellar travel and the side-effects on the lives of the characters.

23 January 2016

The New Yorker: “The Weight of the World”

The Saudis are sitting on a vast reserve of very cheap oil, she continued. Can you blame them for trying to protect that resource and that income for as long as they can? I don’t blame them. It’s very understandable. Let’s do a thought experiment. I come from a country that has only hydro and wind as power resources. If I had been born in a country with fossil-fuel reserves, would I have a different opinion about what’s good for the world? Maybe. Very likely, in fact.

I don’t want to put people into a black box and say, You’re the culprits, and point a blaming finger. It just helps absolutely nothing. Call it my anthropological training. Call it whatever. But I always want to understand: what is behind all of this?

Elizabeth Kolbert

As you may already know, last December a landmark agreement on climate change was signed in Paris. This is a profile (published months before the conference) of the Secretary of the United Nations Framework Convention on Climate Change, Christiana Figueres, the woman leading the efforts to bring every country on Earth together on this major issue for our future. I find her attitude and approach very honest and refreshing – and that’s probably one of the big reasons she managed to pull a deal through this time around. Hopefully she and others with her commitment will remain in charge long enough for this agreement to materialize into concrete measures.

15 January 2016

Medium: “The resolution of the Bitcoin experiment”

Why has Bitcoin failed? It has failed because the community has failed. What was meant to be a new, decentralised form of money that lacked “systemically important institutions” and “too big to fail” has become something even worse: a system completely controlled by just a handful of people. Worse still, the network is on the brink of technical collapse. The mechanisms that should have prevented this outcome have broken down, and as a result there’s no longer much reason to think Bitcoin can actually be better than the existing financial system.

Mike Hearn

A protocol pretending to be a currency while not actually meeting the economical requirements, an experiment shunning proper decision structures in favor of anarchy and short-term interests – is there any wonder it stands at the brink of collapse?

12 January 2016

PLoS Neuroanthropology: “Scott Atran on Youth, Violent Extremism and Promoting Peace”

Last summer, an ICM poll revealed that more than 1 in 4 French youth – of all creeds – between the ages of 18 and 24 have a favorable attitude towards ISIS; and in Barcelona just this month 5 of 11 captured ISIS sympathizers who planned to blow up parts of the city were recent atheist or Christian converts. The unholy alliance of narrow xenophobic nationalism and militant jihad, which play off one another’s fears, are beginning to destabilize the European middle class much as fascism and communism did in the 1920s and 30s, while inciting willingness to sacrifice among both nationalist xenophobes and militant jihadis. By contrast, our own research shows that even among native Western youth, ideals of liberal democracy no longer elicit willingness to make costly sacrifices for their defense.


But the popular notion of a “clash of civilizations” between Islam and the West is woefully misleading. Violent extremism represents not the resurgence of traditional cultures, but their collapse, as young people unmoored from millennial traditions flail about in search of a social identity that gives personal significance and glory. This is the dark side of globalization. They radicalize to find a firm identity in a flattened world: where vertical lines of communication between the generations are replaced by horizontal peer-to-peer attachments that can span the globe. Young people whose grandparents were Stone Age animists in Sulawesi, far removed from the Arab world, told me they dream of fighting in Iraq or Palestine in defense of Islam.

Scott Atran

Wonderful analysis of the extremist phenomenon from its causes to possible ways of converting young people away from these tendencies. It’s very hard to fight people willing to kill (or die) for their beliefs – the best method would be to give them better reasons to live, a positive narrative that feels increasingly absent from Western democracies.

10 January 2016

Future Visions: Original Science Fiction inspired by Microsoft

in Bucharest, Romania

În urmă cu câteva săptămâni a fost lansată o colecție de povestiri mai neobișnuită, prin faptul că a fost comandată și susținută de Microsoft, care a invitat o serie de autori SF în laboratoarele lor de cercetare pentru inspirație despre posibilele tehnologii ale viitorului. De vreme ce este disponibilă gratis, am descărcat‑o pe Kindle și am citit‑o pe finalul de an. Povestirile rezultate, deși se încadrează bine în tema generală a colecției, sunt de o calitate foarte variabilă – aș zice că doar jumătate dintre ele merită citite de sine stătătoare.

Hello, Hello de Seanan McGuire se axează pe tehnologia traducerii în timp real, la care Microsoft lucrează intens pentru Skype, în competiție cu Google. Bazat pe cei mai recenți algoritmi, Paulson dezvoltă un program traducător mai general care ia în considerare inclusiv gesturile și mimica interlocutorului, pentru a putea comunica mai natural cu sora ei Tasha, surdomută din naștere. Spre mirarea ei și iritarea familiei, de la aparatul lui Tasha încep să sosească apeluri anonime – imaginea e înlocuită de un avatar pentru confidențialitate, deci nu poate vedea cine e la celălalt capăt al firului. Pe la jumătate eram destul de convins că știu despre ce e vorba, dar autoarea a reușit să mă surprindă cu o explicație mai bună, așa că o clasific printre povestirile bune. Asta nu înseamnă că nu are minusuri destul de mari, cum ar fi lipsa telefoanelor mobile din peisaj – sora ar putea foarte bine să folosească mesageria mobilă în locul unui aparat fix, ceea ce ar elimina complet suspansul fals. De asemenea, mi‑a fost greu să mi‑o imaginez pe Paulson ca femeie, comportamentul ei pare mai degrabă masculin de‑a lungul poveștii, în special relația cu soția și copiii – dacă o lesbiană gândește la fel cu un bărbat, care mai e scopul diversității?

The Machine Starts de Greg Bear tratează computerele cuantice din perspectiva lui Bose. Echipa din care face parte se află la a treia încercare, ultima șansă de a demonstra viabilitatea proiectului lor. Presați de investitori, programează Bila‑8 cu un set experimental de algoritmi care funcționează perfect din prima, spre ușurarea tuturor. Totuși explicația teoretică pentru care tentativa a reușit atât de bine pare prost înțeleasă și consecințele neprevăzute nu se lasă așteptate. Cea mai hard povestire, dar destul de seacă și prost dezvoltată, plus că consecințele de care vorbeam nu sunt deloc plauzibile dacă iei în considerare legea conservării energiei.

05 January 2016

Nieman Storyboard: “How to Tell Powerful Narratives on Instagram”

So far, this territory has been left to photographers. Since the app’s release in 2010, photojournalists have been using it to great effect, showcasing unpublished images, digging into their archives, sharing ongoing creative projects. Writers, though, have largely stayed away from Instagram as a storytelling platform. There are plenty of reasons for this—it’s not easy, after all, to write short.

But those who wade in will find that storytelling on Instagram is an awesome hack: a purpose for which the thing wasn’t intended, but at which it excels. The app is vibrant, flexible, and unusually transcendent. In mobile terms, it’s immediately more democratic than filter-heavy Facebook, less terse than Twitter, less ephemeral than Snapchat. And, most important to me, is the platform’s reach: Instagram provides a creative space where voices and views that might otherwise be ignored, lost, or mangled during their brush with journalism can be shared, beautifully, with almost anyone.

Neil Shea

A contributor to National Geographic Magazine and other publications, Neil Shea recounts how he discovered Instagram as the perfect medium to share stories from his journalistic projects, small glimpses into the lives of the people he meets that would otherwise get lost in the final, more formal piece. It’s one of the fascinating ways people are using the Facebook-owned app and its flexibility helps make it more and more popular. Although I’m not entirely convinced many people actually read the stories accompanying photos, this experience is a valid reason for viewing Instagram as a future platform for news, but I think it will develop as complementary to traditional news stories rather than replacing them.

04 January 2016

The New Yorker: “In Silicon Valley now, it’s almost always Winner Takes All”

As someone who has felt, first-hand, the agony of shuttering the doors of his startup, I feel Paul’s pain. But I want to focus on what Branson, a self-made billionaire, who is more often right than wrong, said about ride-sharing not being a “winner-takes-all” market. What Branson says is generally true for companies that sell analog products, such as packaged goods or soda, or analog services, such as air travel. Coke isn’t going to drive Pepsi out of business, and Toyota isn’t going to eliminate Honda. But in today’s Internet-always-on world, that maxim increasingly doesn’t hold true. Most competition in Silicon Valley now heads toward there being one monopolistic winner. And that is why it is hard not to see that, right now, the only competition that matters in ride-sharing is between the two largest companies: Uber and Lyft.

In the course of nearly two decades of closely following (and writing about) Silicon Valley, I have seen products and markets go through three distinct phases. The first is when there is a new idea, product, service, or technology dreamed up by a clever person or group of people. For a brief while, that idea becomes popular, which leads to the emergence of dozens of imitators, funded in part by the venture community. Most of these companies die. When the dust settles, there are one or two or three players left standing. Rarely do you end up with true competition.

Om Malik

While the conclusion sounds true at first, especially after the news of one competitor closing doors at the end of 2015, I think it’s a bit early to call win for the ride sharing market. Every example mentioned in the post as dominant, monopolistic winner is a mature public company, with a steady flow on revenues and profit – none of the startups operating in ride-sharing are at this point yet. The way I see it, Uber and co. do qualify as ‘analog’ services more than digital: while there’s an algorithm behind it optimizing rides and fares, the actual transaction with the customer involves a person driving a car, not on-screen interactions like Google, Facebook and Amazon. The real point of failure for Uber is the driver, and, if repeated reports are to be believed, drivers are not happy with the company, so Uber has high attrition and this could hurt customer satisfaction in the long run. 

03 January 2016

Reading stats for 2015

As 2015 has come to an end, it’s time for another short look at my reading for the past year – courtesy of Goodreads. At first look, I shelved even less books than in 2014 (well, one less). The page count tells a different story though, with Goodreads reporting around 7600 pages read in 2015 – not counting the short story collection I started in December and not finished yet – about 830 pages more than in 2014, but still less than the peak in 2013. The higher page count reflects that over the past year I preferred longer novels to short stories, the largest being Anathem and the Prince of Nothing series, each counting thousands of pages. I am usually conservative with my ratings, and this year was no different: I ended up rating most books I finished 3 or 4 stars. Another thing I noticed is that I am rarely impressed with new releases: I bought both Something coming through and Binti as pre-releases on Amazon and both ended up in my 2-stars shelf. So I think for the future I will wait until a couple of reviews are published before jumping into new territory.

My Goodreads reading stats for 2015