Thought starters: Artificial Intelligence and its malcontents

Karen Hao takes a look at how OpenAI is looking to develop artificial general intelligence and the conflicts between getting there first and fulfilling its founding ethical principles don’t always make for perfect bedfellows:

“They are using sophisticated technical practices to try to answer social problems with AI,” echoes Britt Paris of Rutgers. “It seems like they don’t really have the capabilities to actually understand the social. They just understand that that’s a sort of a lucrative place to be positioning themselves right now.”

Whilst were on the subject of artificial intelligence, it’s worth reading Paul Grimstad’s profile of Alan Turing whose life came to a premature end due to homophobia of the post war era:

“It is fortifying to remember that the very idea of artificial intelligence was conceived by one of the more unquantifiably original minds of the twentieth century. It is hard to imagine a computer being able to do what Alan Turing did.”

Ian Parker wrote a great profile of Yuval Noah Harari who I’ve been a big fan of since reading Sapiens. It’s well worth a read but did leave me feeling rather wary of his assiduous avoidance of taking a stance on the issues he writes about.

The concept of filter bubbles has provided a tidy justification for what many of us see as the growing polarisation in this era of social media. Research from the Dr Richard Fletcher at the Reuters Institute provides a much more nuanced analysis in a world where there’s a growing abundance of news sources and algorithms aren’t the only means of discovery:

“Most of the best available independent empirical evidence seems to suggest that online news use on search and social media is more diverse. But there’s a possibility that this diversity is causing some kind of polarisation, in both attitudes and usage. This is interesting, because in some ways it’s the opposite of what the filter bubble hypothesis predicted.”

Tyler Cowen’s Marginal Revolutions blog is a regular source of nuggets of information. The highlighting of research from Jason L Cummings provides an interesting look at a possible driver for the rise of Donald Trump

“Black women for instance, present a consistent pattern of improvement in happiness across decades, while White women display a persistent pattern of decline. In contrast, Black men experienced a discernable pattern of improvement in happiness between the 1970s and 1990s, followed by a leveling off in the early-2000s. White men experienced moderate gains in happiness between the 1970s and 1990s, but after the Great Recession/Obama Era, White male happiness followed a pattern of unprecedented decline, with the “happiness advantage” they once enjoyed (as a group) over Black men and women largely vanishing.”

Peter Thiel’s often quoted statement We wanted flying cars, instead we got 140 characters has long given people food for thought particularly as people point to the lack of growth in productivity in the global economy.  The It’s Only Chemo blog provides a somewhat different point of view as we all become infovores less obsessed with the material, something not necessarily picked up in traditional measures:

“Perhaps much of this is explained by the Alchian-Allen Theorem. There is so much to be gained by simply sitting at your screen and surfing, exploring the cultural niches of YouTube or learning Game Theory online or simply playing videos games. We haven’t yet realised that our minds are the new frontier. And therefore the returns to any sort of physical world accomplishments are much diminished.”

For a look at traffic modelling, this video is mesmerising. What I also find interesting is what it leaves out such as the costs of different options and the impact the different solutions have on people who aren’t confident motorists (elderly motorists, cyclists, pedestrians etc). What you exclude sometimes says more about you than what you include:

 Culture

I find winter time is a great opportunity to catch films pre/post award seasons or simply catch those films that missed your attention when they came out on the silver screen. I’ve been really enjoying using Letterboxd to track films I’ve seen (apparently 741 films seen at last count with 335 films on my watch list) and you can catch me here.

Bong Joon-Ho’s Parasite is deserving of the praise that’s been piled on it in my opinion and explainer videos from Thomas Flight and Nerdwriter give an idea of the level thought that has gone into the finished film.

I appreciated 1917 nail biting trip the trenches but it was Terrence Malick’s A Hidden Life that made the bigger impression. The film is a beautiful look at the struggles endured by an Austrian conscientious objector and his family during World War II.

Whilst were on World War II, Ari Richter’s illustrated account of his trip to Auschwitz provides a valuable look at how histories are being rewritten by populist governments to serve their own ends.

I can remember Cerith Wyn Evans’ work catching my eye at the Tate Britain and it’s great to see him have the whole of White Cube Bermondsey to explore his artistic vision. Well worth a visit.

Tools

Aegisub: Whilst the open source software’s website looks like it’s something from an earlier era, the tool for creating subtitles hits the spot if you’re posting videos to Twitter or other video platforms. I used this for our recent Bobby Seagull video for our petition to end library austerity.

DIY Captions Launcher for Youtube: Transcribing video is a painful task I wouldn’t wish on anyone. Google has been doing great work with its subtitling technology and this Chrome Extension helps you pull down transcripts. There’s inevitably going to be some corrections involved but it does much of the leg work for you. In my case, this has been great for transcribing videos from CILIP Conference’s last year, I task I fear would otherwise never have been completed.