New Week #103
Google are building the all-purpose helper robot of our dreams. Plus more news and analysis from this week.
Welcome to the mid-week update from New World Same Humans, a newsletter on trends, technology, and society by David Mattin.
If you’re reading this and haven’t yet subscribed, join 22,000+ curious souls on a journey to build a better future 🚀🔮
It’s been a mad seven days on the work front, and that means a truncated and text-only instalment of the newsletter this week.
Still, I want to share two quick updates on long-running NWSH obsessions.
First, Google unveils new research on their quest to create robots we can talk to. And second, an innovative designer offers a glimpse of the AI-fuelled future of creative work.
I’ll be back next week with a bumper edition, complete with audio version as usual; promise! For now, let’s get into it.
🤖 Talk it out
Back in New Week #95 I wrote on Google’s ongoing work to create everyday helper robots that can understand us.
That instalment focused on PaLM-SayCan, a system that seeks to marry Google’s massive AI language model, PaLM, with the hardware taking shape inside its Everyday Robots division.
This week, further news from Google Robotics. A new paper offered yet another glimpse of their work to create ‘interactive, real-time, natural language-instructable robots in the real world.’
The research team trained an AI on hundreds of thousands of ‘language-annotated trajectories’ — labelled videos of a robotic arm attempting to perform specific block rearrangement tasks in a defined space. The result is an AI-fuelled arm that can respond to ordinary spoken instructions such as move the green star between the yellow blocks, or nudge the blue triangle down a bit.
The Google team estimate their arm can achieve a 94% success rate on a set of 87,000 unique natural language commands. They say they’ll soon release both the code and the dataset, Language Table, which is comprised of over 600,000 labelled videos.
⚡ NWSH Take: It might feel easy to shrug your shoulders here. So this robot arm can shuffle blocks around; big deal. It’s hardly Midjourney! But think again: this kind of work is ushering slowly but steadily towards a universal helper-bot; the kind of robot 20th-century futurists loved to assure us was just around the corner (fans of retro futurism, you must click that link). // Such robots will come first to industry; just imagine an Amazon fulfilment centre filled with robots the workers can talk to as easily as they can one another. But soon enough, they’ll come to our homes, too. It’s time I wrote more on all this, but to catch up with the NWSH take read one of this newsletter’s anchor essays, Our Coming Robot Utopia.
🎨 Design genie
I’ve been writing regularly on the incredible advances being made by AI generative models. Recently that meant news from Google (yet more Google!) on its incredible new text-to-video tools, Imagen and Phenaki.
We’re amid a revolution that’s seeing AI and human creativity collide, and NWSH is obsessed.
This week, one Twitter user offered us a glimpse of what that collision may come to look like in practice. He hacked together an AI-fuelled design assistant that turns spoken ideas into amazing images:
The tech stack here? This designer used The MetaHuman Creator from Epic Games to create the assistant’s avatar. OpenAI’s Whisper handles the speech recognition, and GPT-3, also from OpenAI, is the model that’s fuelling language competence. Finally, the text to image model Stable Diffusion generates the illustrations.
Put it all together and the result is something close to magic. A virtual human that can generate and refine any image at your whim, in seconds.
⚡ NWSH Take: Taken alone, none of this is new; NWSH has covered the MetaHuman Creator, GPT-3, and Stable Diffusion extensively across the last 12 months. But tie all those tools together, and use Whisper to create a seamless UX, and we get a glimpse of fast-approaching future. In short, this is what the future of work could look like for designers and a wide range of other creative and knowledge professionals, from architects, to web developers, to writers. Those workers will constantly spar with and iterate the work of AI-fuelled assistants that can generate endless drafts and revisions. It turns every creative professional into the boss of their own hyper-productive output factory. Want a glimpse of the work we can expect? Take a look at this animated anime music video, created in under 12 hours using Stable Diffusion. // The upshot? Standard issue ‘good design’ and many other kinds of creative output — from copywriting to code — will be commoditised. Top tier creatives will need to learn to pilot generative models better than the Average Joe. AI whispering is about to become a whole lot more important.
🗓️ Also this week
❄️ New research suggests the next pandemic may be unleashed by melting Arctic ice. An analysis of ice from Lake Hazen in the Arctic suggests it is host to ancient viruses and bacteria that could infect local wildlife if the ice melts.
😱 The British Broadcasting Corporation is preparing secret scripts for emergency broadcasts in case of energy blackouts this winter. The UK’s National Grid, which oversees energy supply, has warned that worst case scenarios may see the country’s electricity supplies may falter in the coming months.
🦾 The Netherlands has become the first NATO country to trial armed military robots. Four armed unmanned ground vehicles — essentially mini-tanks — were deployed in a trial in Lithuania. Back in New Week #101 I wrote about iconic robotics startup Boston Dynamics, and their promise not to weaponise their devices.
🚚 Ikea is trialling driverless truck deliveries in Texas. The furniture giant is partnering with driverless technology startup Kodiak; since August a driverless truck has been moving furniture from an Ikea distribution centre near Houston to a retail store near Dallas.
👨💻 The US Bank of Federal Reserve says remote workers sleep more and work less than they used to when they went to an office. New data shows ‘a substantial fall in time spent working’ and an increase in time spent on leisure. Back in New Week #100 I wrote about data from the Work Trend Index report that shows managers are worried about their remote teams slacking off at home.
⚽️ The government of Qatar says all visitors to this year’s World Cup must download and run two apps during their stay. One of the apps, Ehteraz, is intended to help track covid-19, but experts say it can also read, edit, and delete all content on a phone, and prevent the phone from being put on airplane mode.
🌍 Humans of Earth
Key metrics to help you keep track of Project Human.
🙋 Global population: 7,982,632,794
🌊 Earths currently needed: 1.7910270789
💉 Global population vaccinated: 63.0%
🗓️ 2022 progress bar: 77% complete
📖 On this day: On 21 October 1879 Thomas Edison files for a patent for his design of the incandescent light bulb.
Future Autonomous
Thanks for reading this week.
The ongoing collision between human creativity and machine intelligence is yet another classic case of new world, same humans.
This newsletter will keep working to make sense of what’s happening, and what it means for our shared future.
And there’s one thing you can do to help: share!
Now you’ve reached the end of this week’s instalment, why not forward the email to someone who’d also enjoy it? Or share it across one of your social networks, with a note on why you found it valuable. Remember: the larger and more diverse the NWSH community becomes, the better for all of us.
I’ll be back next week. Until then, be well,
David.
P.S Huge thanks to Nikki Ritmeijer for the illustration at the top of this email. And to Monique van Dusseldorp for additional research and analysis.