Predicting Technology Impact; Tesla’s Secret Batteries; Computers and Common Sense; Renewable Energy poised to eclipse coal in the US?

I almost had forgotten about it, but when I attended the House of Beautiful Business (HoBB) last November in Lisbon, on top of holding a session on what poetry brings to business, I took part in a sort of experiment organized by HoBB in collaboration with Porsche. 

I agreed to get in a room with Amy Whitaker, professor at NYU and author of the book “Art Thinking” and have a discussion on the power of art and its influence on business. The conversation would be recorded and possibly turned into a podcast. Important detail, Amy and I met for the first time at the HoBB and we never discussed in advance what should be part of the discussion, i.e. the conversation would be completely improvised, no agenda or story to follow. 

Eventually, that conversation made to a podcast, and is now part of the podcast series by Porsche, and it launched this week. It was interesting listening to it, as the discussion took place only six months ago, but it is such a different world now, so I was wondering, will it be still relevant in this new context?

I was very surprised to see how the conversation is still relevant in today’s context, and actually has even increased its relevance. With Amy we discussed the importance of Art Thinking, how poetry can contribute to make sense of a business environment which is increasingly volatile and uncertain, how business should focus on the senses, i.e. on how things make us feel, as this (feeling) is one of the things that make us human, and making business more human means making it more beautiful. We also discussed how, in a world where the analytical part is being taken over, more and more, by technology, imagination, a true human characteristic, is going to become more and more important and art and creativity are essential to it. If any of the above intrigued you, you might want to listen to the podcast here(apple) and here (spotify).

P.S. Apologies to those of you received this also as a post. LinkedIn publishing software and I have not been getting along well lately...

Some highlights from this week's edition include:

●      Some technologies start out as "toys" but end up changing the world - the airplane, the personal computer, the smartphone. Others remain toys and never break into the mainstream. Benedict Evans explores how to tell these apart.

●      The world's stigmas around data collection, AI, and labor-replacing robots have been quietly put into a corner during the pandemic. That's not exactly a good thing, as we learn from two Georgia Tech professors.

●      Social distancing and stay-at-home orders are sending more autonomous delivery robots to the streets, but these aren’t the self-driving trucks and robotaxis that automakers have promised to deploy for years - they are remotely guided robots that operate on sidewalks.

●      Tesla is set to launch a new, low-cost EV battery that will see it sell electric vehicles at the price of gasoline cars - or lower. The company will use its new “million-mile” battery in the Model 3 sedan in China this year or next, and later introduce it to other markets.

●      Avantium, a Dutch biochemical company, plans to turn plants into plastic to help reduce the reliance on fossil-based materials. The move has already won the support of beverage makers such as Carlsberg, Danone, and Coca-Cola

●      Helen Papagiannis, founder at XR Goes Pop and author of Augmented Human, guest posts on WEF on augmented reality's potential during the pandemic. Papagiannis sees three things AR doing very well: visualization, annotation, and storytelling.

...along with a number of other enlightening reads. 

Human and Machine

Common Sense Comes Closer to Computers

It's easy to mistake the surprisingly realistic - but ultimately shallow - results by OpenAI's GPT-2 model for intelligence, as anyone who's played with talktotransformer.com or AI Dungeon can attest. But it turns out to be a Clever Hanseffect, as demonstrated by AI researcher Gary Marcus, who asked GPT-2 what happens when you drop a match onto a stack of kindling. The answer, as you can test for yourself on Talk to Transformer, never makes sense, and that's because computers lack common sense - the "dark matter of AI."

Algorithms don't read between the lines. Hand-crafted approaches are tenuous, due to the ambiguity of language and the fickleness of hard-coded relations. Deep learning hasn't fared well either, as seen with GPT-2, but that hasn't stopped researchers from building common sense systems that take a middle approach. Commonsense transformers- or COMET - opts for generating plausible yet imperfect responses to new input, rather than using deterministic models (see the image below). You can try COMET here.

Not Even Wrong: Ways to Predict Tech

"A lot of really important technologies started out looking like expensive, impractical toys.... aircraft, cars, telephones, mobile phones and personal computers were all dismissed." - Benedict Evans

Dismissing an important early technology as a "toy", like in the statement above, is not indicative that it will go on to change the world. It might be just that - a useless toy. But the toy statement is vapid to begin with. It's "not only not right; it is not even wrong," to quote theoretical physicist Wolfgang Pauli. Evans says we need to go deeper, asking, "do we have a theory for why this will get better, or why it won't, and for why people will change their behaviour, or for why they won't?"

To illustrate the point, Evans compares the Wright Flier to the Bell Rocket Belt - both "expensive impractical toys," but one went on to change history. The difference? The Flier had a roadmap: "There was no reason why [the Flier] couldn't get much better, very quickly.... There was a very clear and obvious path to make it better. Conversely, the Rocket Belt flew for 21 seconds because it used almost a litre of fuel per second - to fly like this for half a hour you'd need almost two tonnes of fuel.... There was no roadmap to make it better without changing the laws of physics. We don’t just know that now - we knew it in 1962."

Creativity

This AI-Generated Dictionary Is Very Cool and Also Terrifying.

"Patiefarge" is a noun meaning "a group of four dogs," and it's completely made up. The word was generated by This Word Does Not Exist, a project by former Instagram developer Thomas Dimson, who trained a GPT-2 language model using the Oxford English Dictionary. A new word is generated every time you refresh the site.

It's Cool to Look Terrifying on Pandemic Instagram

"All day long, I flip through Instagram stories, and watch one augmented diary of life inside after another."

With Instagram backdrops downgraded from breathtaking vistas to drab walls, Augmented Reality creators are making something worth sharing, whether it's a mouse dressed as the Pope dancing on a kitchen floor or an e-makeup artists creating floral masks for 3D influencers. Atlantic's Kaitlyn Tiffany speaks to some of Instagram's talented AR artists, like Mitsuko Ono and David O'Reilly, and director of the XReality Center at the New School Maya Georgieva.

Georgieva notes that "during lockdown, she’s seen filters getting both wilder and even more popular, because 'you can transform instantly.' AR is so mainstream now... that it's even showing up in schools and workplaces. Teachers might still raise an eyebrow at a weird Zoom background that covers a student's entire face, but they probably accept at this point that some kids are going to call in with a background from [SpongeBob]."

Life Sciences

AI, Robots, and Ethics in the Age of COVID-19

"More of us are overlooking our former uneasiness about robots and AI when the technology's perceived value outweighs its anticipated downsides. But there are dangers to this newfound embrace."

The western world's stigmas around data collection, AI, and labor-replacing robots have been quietly put into a corner during the pandemic. Robotic telepresence platforms in Japan stand in for students, chatbots are helping alleviate the influx of user questions on online education platforms and at companies like PayPal, and recycling bots are being tested at sorting facilities, not to mention the increased use of healthtech like telehealth platforms and surveillance drones.

Chair of the School of Interactive Computing at Georgia Tech Ayanna Howard and director of the graduate research ethics programs Jason Borenstein suggest several critical considerations organizations should make before they become more reliant on tech like AI and robotics: "opportunities for bias that we know exist in AI are still a concern," along with the "privacy concerns with respect to data collection and data accuracy."

Spaces

Help Wanted: Autonomous Robot Guide

Social distancing and stay-at-home orders are sending more autonomous delivery robots to the streets. These aren’t the fully autonomous vehicles that automakers have promised to deploy for years; they're remotely guided robots that operate on sidewalks, controlled by human guides who monitor and direct their movements in real-time.

Postmates, for example, operates sidewalk delivery robots in San Francisco and Los Angeles via a partnership with Phantom Auto. The uptick in demand for contactless delivery has led the company to increase its delivery robots by 30% since March 17. Other autonomous delivery startups such as Refraction AI and Starship Technologies have also seen increased demand.

See China's COVID-19 Contact Tracing System in Action

Michael Ryan, WHO's emergencies chief, has likened the reopening of pandemic-stricken areas to "driving a car with eyes closed." Countries around the world are racing to develop contact tracing apps to keep track of infected patients and mitigate the spread of the virus; in some countries, privacy considerations are stagnating developments. Earlier this month, health experts warned that Google and Apple’s tracking systems could be "practically useless" if the companies continued to impose strict privacy rules.

Unburdened by the privacy debate, China has already rolled out its color-coded QR contact tracing system in over 100 cities. The system assigns residents color codes that indicate whether they should be in self-isolation, supervised quarantine, or are free to move around. In Europe, authorities are exploring the use of decentralized systems to address privacy concerns:

"In these decentralized systems, information on a person's movements is left on their device unless they test positive for COVID-19, unlike 'centralized' systems where every morsel of data ends up on a central server."

Materials

In a First, Renewable Energy Is Poised to Eclipse Coal in U.S.

The pandemic is accelerating the decline of coal and pushing the US closer to renewable energy. While parties attempt to keep the mining industry afloat, renewables are expected to pull ahead of coal this year. Stay-at-home orders have reduced the demand for electricity, and coal-fueled energy was the first to be cut by utility companies due to high costs.

This decade has seen a series of coal plant closures due to clean energy regulations and reduced prices of renewables like wind and natural gas. Despite these developments, Michelle Bloodworth, CEO at America's Power, says this isn't the end for coal: "There is still a significant amount of coal that's going to be needed in the future to make sure we don't risk and threaten the reliability of the grid."

The End of Plastic? New Plant-Based Bottles Will Degrade in a Year

Avantium, a Dutch biochemical company, plans to turn plants into plastic to help reduce the reliance on fossil-based materials. The move has already won the support of beverage makers such as Carlsberg, Danone, and Coca-Cola, and the Avantium plans to reveal more partnerships with food and drink companies later this summer.

The company's process involves extracting sugars from biowaste and sustainable plants such as corn and wheat to make recyclable plastics that degrade much faster than fossil-based plastics. The material will be resilient enough to hold carbonated drinks, and will degrade after a year. Avantium could start distributing the new plastics by 2023.

Processors

Tesla's Secret Batteries Aim to Rework the Math for Electric Cars and the Grid

Tesla is set to launch a new, low-cost EV battery that will see if it can sell electric vehicles at the price of gasoline cars - or lower. The company will use its new “million-mile” battery in the Model 3 sedan in China this year or next, and later introduce it to other markets. Tesla has developed the battery in collaboration with China’s Contemporary Amperex Technology company.

To increase the life of its EV batteries, the company is using chemical additives, materials and coatings that reduce internal stress, and technologies such as low-cobalt and cobalt-free battery chemistries. The company will achieve the low-price target by leveraging high-speed, automated manufacturing processes to cut on labor costs, and with the help of CATL’s cell-to-pack technology, a cheaper and simpler way of packaging battery cells.

Interfaces

Spatial Goes Free, Aiming to Become the Zoom of Virtual Collaboration

Launched in 2016, Spatial has been focusing on the corporate market for the past few years. As offices continue to operate remotely, the company is extending its reach by including a free tier for its collaboration platform, and adding support for browsers, smartphones, and the Oculus Quest VR headset. Engadget details their experience in a Spatial session with CEO Anand Agarawala.

3 Ways Augmented Reality Can Have a Positive Impact on Society

Helen Papagiannis, founder at XR Goes Pop and author of Augmented Human, guest posts on WEF on augmented reality's potential during the pandemic. Papagiannis sees three things AR doing very well:

●      Visualization: "All 185 first-year medical students at Case Western Reserve University are using HoloLens and HoloAnatomy to learn from their own homes. HoloAnatomy helps students learn about the human body in ways not otherwise possible."

●      Annotation: "London's National Theatre is using AR to help make its performances more accessible for people who are deaf and hard of hearing. When wearing a pair of smart caption glasses, users see a transcript of the dialogue and descriptions of the sound from a performance displayed on the lenses."

●      Storytelling: "Lessons in Herstory uses AR to help rewrite history books in the classroom and inspire the leaders of tomorrow by featuring stories of powerful women."