Here’s what happened in the world of artificial intelligence in 2020

The year 2020 was long and treacherous, but the biggest bright spot for me was the official launch of Neural . That’s our AI sub-brand here at TNW and the section you’re reading this article in.

More specifically, Neural is me (Tristan Greene), Thomas Macaulay, Ivan Mehta, and the contributors and colleagues who help us put out fresh, original, exciting content in the world of machine learning every day.

It was a tough year to be a reporter but Thomas and Ivan managed to exceed our expectations at every turn with incredible insight and consistent excellence. With that in mind, I’m proud to present some of my favorite articles from Tom and Ivan this year.

Between the two of them they covered some of the biggest events, breakthroughs, and stories in the world of machine learning and artificial intelligence. But, more importantly, they provided keen insight and analysis that you won’t find anywhere else. And they also adhered to our biggest principal here at Neural: we cover AI for humans (not robots, businesses, or governments).

So, if you’ll indulge me, here’s an “Editor’s Choice” list of just a few of the many articles my team published this year.

But first, here’s my contribution:

Stories pictured above here and here .

Facebook is reportedly developing AI to summarize news — what could go wrong?

TL;DR: This AI summarizes research papers so you don’t have to

This AI jacked Eminem’s flow to make a Zuckerberg diss track

AI mistakes referee’s bald head for football — hilarity ensued

Google’s new machine learning tool turns your awful humming into a beautiful violin solo

India finally commits to quantum computing, promises $1.12B investment

MIT researchers developed a text-based system that tricks Google’s AI

This AI model improves and colorizes old videos and the results are stunning

MIT and Microsoft build a new algorithm to find hidden links between historic paintings

How Facebook’s Yann LeCun is charting a path to human-level artificial intelligence

Oh great, the EU has ditched its facial recognition ban

New Zealand’s first AI police officer reports for duty

AI interviewer asks Billie Eilish the weirdest questions

Why AI systems should be recognized as inventors

AI Jesus will help you embrace the apocalypse

Meet the AI-generated bipartisan president of the US

Someone let a GPT-3 bot loose on Reddit — it didn’t end well

Masks won’t protect you from facial recognition

Backlash intensifies: Pichai’s promises do little to quell outrage over Timnit Gebru’s firing

Honey traps and bribery: Ex-Cambridge Analytica CEO slapped with 7-year directorship ban

AI discovers that every lion has a unique and trackable roar

And those are just a small sample of the wonderful work we’ve put out here at Neural . Check back in with us in 2021 where we’ll continue to bring you news, analysis, and trusted opinions on the world of machine learning and its impact on humans.

Google’s new crowd predictions for Maps may coax me back onto subways

Subways have never been my favorite mode of transit, but the onset of the pandemic made them even less appealing. Sixteen months later, Google Maps may finally have found a way to lure me back underground.

The service is now piloting the option to see live data on the crowdedness of individual train carriages.

Up front : Google initially launched transit crowdedness predictions in 2019 , when people were more worried about finding a seat than catching a potentially fatal infection.

The feature uses a combination of AI , historical location trends, and user contributions to estimate the seating and standing capacity on buses, trains, and subways.

The predictions have now been expanded from nearly 200 cities to more than 10,000 transit agencies in 100 countries. In addition, Google is now piloting the ability to see live crowdedness information on individual carriages, so you can choose exactly where you want to hop on.

The new data is currently available in New York and Sydney. Google says more cities are coming soon.

Quick take: The drive to get people back into offices is making many public transport systems busy again. Crowdedness predictions can make the experience less painful, whether you’re worried about COVID or just hate jam-packed carriages.

Now I only need Google to sort out the subway‘s sweltering heat, high costs, constant delays, and horrible commuters, and I’ll gladly swap my bike for a trip underground.

Greetings Humanoids! Did you know we have a newsletter all about AI? You can subscribe to it right here .

‘Slaughterbots’ are a step away from your neighborhood — and we need a ban

If you’re concerned about killer robots in the military, brace yourself for their arrival in civilian hands.

AI weapons will soon spread from battlefields to city streets, campaigners have warned, unless new global rules are imposed on the tech.

A UN conference this week can alleviate the fears . In a  Geneva meeting that starts today, delegates will debate banning weapons that target people without “meaningful human control.”

Activists, however, are pessimistic about the outcome. While some countries have endorsed new laws, the world’s leading military powers don’t appear enthusiastic.

The US, for instance, has rebuffed calls to regulate lethal autonomous weapons (LAWs). Instead, American officials have proposed developing “a non-binding code of conduct.” This may outline some principles of use, but there would be no legal obligations to abide by them.

China and Russia have also shown little appetite for a global treaty. As the UN meeting requires a unanimous agreement on any new rules, the prospects of meaningful restrictions look bleak.

Ultimately, the opposition to new rules may prove self-destructive.

“The biggest losers from this are going to be countries that are militarily dominant, because these weapons are incredibly cheap,” Max Tegmark, an AI researcher at MIT and co-founder of the Future of Life Insitute (FLI) , told TNW.

“They’ll be small, cheap and light like smartphones, and incredibly versatile and powerful. It’s clearly not in the national security interest of these countries to legalize super-powerful weapons of mass destruction.”

The FLI has produced a film that envisions the future of these unregulated “slaughterbots.”

The images are nightmarish. They depict mass executions at polling stations, killer robots used in heists, and drones with facial recognition targeting both individuals and specific demographic groups.

This dystopian vision could one day become a reality. LAWs may have already killed soldiers without gaining permission from a human operator. In the future, Tegmark expects most of the weapons to be used by and on civilians:

While Tegmark doesn’t expect a global ban to emerge this week, he believes individual nations may gradually sign up to new rules. In time, he hopes LAWs become so stigmatized that every military power is pressured to agree on prohibitions. If they don’t, AI weapons of mass destruction may enter the mainstream.

Leave A Comment