New Week Same Humans #35
Elon Musk built a school for his children and now it's gone global. Scientists teach an AI to recognise sarcasm. 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 you haven’t yet subscribed, then join 16,000+ curious souls on a journey to build a better shared future 🚀🔮
💡 On Sunday I wrote about satellites, the night sky, and global challenges in the 21st-century. Go here to read Do We Need a World Government? 💡
This week, the tiny SpaceX school that taught Elon Musk’s children says it wants to reimagine education for the 21st-century.
Also, an AI was taught to understand sarcasm. And Google has a cunning plan to answer your search queries without sending you to external sites.
Plus, a new technique will allow scientists to simulate complex universes in a fraction of the time it used to take.
Let’s go!
🧑🏫 Synthesising young minds
Want to glimpse the future of education? EdTech startup Synthesis announced a $5 million funding round this week.
Synthesis has a fascinating origin story. Back in 2014, Elon Musk invited a young teacher to found a small private school on the SpaceX campus in California. He instructed that teacher, Josh Dahn, to reinvent education for the 21st-century and roll out a sparkling curriculum for his children.
Now, Dahn has turned the programme he devised into an online offering. For $180 a month ambitious parents can enrol children aged eight to 14 in a one-hour per week ‘enrichment programme’, which teaches pupils to be ‘enthralled by complexity and solving for the unknown.’
I’ll let Synthesis users tell you more:
You get the idea: Elon Musk’s children did this, now yours can too!
Synthesis launched three months ago; they say subscription revenue is already at an annual run rate of $3 million.
⚡ NWSH Take: Last year we saw a connected world collide with the pandemic to catalyse the disruption of higher education. Schooling is a more difficult proposition for innovators. But Synthesis is one attempt at an answer. // As you’d expect, the startup is scathing about traditional schools: too much rote-learning, too many tests. It has a point: mainstream schools have failed to adapt around the 21st-century knowledge economy, and we do need more emphasis on problem-solving and creative collaboration. // But Synthesis is also a gloriously baroque pageant of the Silicon Valley worldview – embrace failure! be an innovator! you win or you learn! – and an unwitting guide to what the global techno-financial elite want for their children. Status-seeking, and immersion in a high-prestige social code, is as visible here as in any Harvard admissions interview. // In short, two reasons to keep watching. First, can Synthesis and similar platforms meaningfully challenge traditional schooling? Second, this startup is an evolving indicator of what the 21st-century global elite value. Where they go, the rest of us will tend to follow.
🤖 Language models, yeah right
Researchers at the US Defense Advanced Research Projects Agency (DARPA) say they’ve created an AI deep learning model that can detect sarcasm. Brilliant work guys, well done.
No, seriously, that is good work. Sentiment, including sarcasm, is a hugely tough problem for AI language models. Scientists at DARPA trained a model to achieve a ‘nearly perfect score’ for identifying sarcasm when pointed at a data set drawn from Twitter.
This work takes us a step closer to natural conversation with AIs. And this week also delivered a signal of just how important those conversations will become.
Researchers at Google published a paper that envisages an entirely new form of online search. One that leverages a deep learning language model – essentially a more advanced version of GPT-3 – to understand a search query and deliver a full answer in natural language, rather than a list of search results.
That service would fulfil Google’s ultimate dream: handle queries without sending users away from the platform. Earlier this year the company announced the development of a new language model with 1 trillion parameters; GPT-3 has 175 billion.
⚡ NWSH Take: Pretty soon the world will be full of digital entities – including fully-realised virtual humans – that want to strike up a conversation. They’ll be chatting to you inside video games, dealing with customer service issues, luring you to social platforms, and trying to sell you stuff. And then there’s the search revolution that Google has cooking. // It’s a new frontier full of possibility. And ethical danger, too. One example? This week, choose your own adventure video game AI Dungeon, which is fuelled by GPT-3, turned on a new moderating system after it was discovered that some players were typing in prompts that caused the game to generate horrific text content. // Ethics research, then, is paramount. Google’s AI ethics team disintegrated earlier this year amid claims of a culture of prejudice and indifference; now the tech giant says it will double the size of the team. // The truth that’s rarely spoken? It’s unlikely we’ll ever create AIs that carry a zero possibility of saying anything offensive. After all, we’ve never managed to create a human like that.
⚙️ A toolkit for world builders
Nintendo launched Game Builder Garage this week.
It’s a kind of Lego kit for video game building, and allows the user to drag, drop, and click their way to building their own racing game, alien blaster, and more.
I’ve written before on how young gamers are building the next version of the internet. That version is shared, social, metaversal worlds, and children are learning how to create them inside video games such as Roblox. Now, add Game Builder Garage to the mix.
If the next iteration of the internet is the metaverse, then world builders are to the next decade what app developers were to the 2010s. Expect a special report on this subject soon!
🏃 Quantify this
Last week I mentioned news that the next Apple Watch will be able to monitor blood glucose and alcohol levels.
That’s thanks to innovative sensors made by UK company Rockley Photonics; those sensors beam infra-red light through the skin.
This week comes news that researchers at Columbia University have created a single-chip system so small it can be injected into the human body. The chip, no bigger than a dust mite, could be used to monitor internal body temperature, glucose levels, respiratory function, and more.
⚡ NWSH Take: Back in the early 2010s they used to talk about the quantified self at TED and in the pages of Wired. Then it just became daily life. A great example of how fringe behaviours turn mainstream. // But the often overlooked part of this story? Fitbits, Runkeeper, and the Apple Watch were only the first manifestation of this trend. Now, a new chapter is beginning. One fuelled by next-generation sensors that empower ordinary users to track metrics that currently aren’t even available in hospitals. Want to monitor real-time blood glucose levels across the day and see how they correlate to energy? Now, you can. // Expect new diet, exercise, and lifestyle practises to emerge out of the Cyborgian Self. As ever, they’ll be driven by age-old human needs: wellness, status, meaning, and more.
🗓️ Also this week
✋ Germany has banned Facebook from using WhatsApp customer data. The regulator says the move is intended to ‘safeguard the rights and freedoms’ of German users. Facebook say they will appeal.
🌬️ The US has approved its first major offshore wind farm. The farm will be 14 miles off the coast of Massachusetts, and will create enough energy to power 400,000 homes.
🤖 Researchers in the Netherlands created a group of autonomous, self-learning robots. The robots leverage a simple algorithm to adapt their group behaviour to changing terrain.
🏘️ Amazon says it will soon launch its Sidewalk ecosystem of connected devices. Sidewalk connects Amazon Echo speakers, Ring devices including doorbells and more, to create a neighbourhood network.
💸 An NFT collection of nine digital characters, called CryptoPunks, sold for $16.9 million. CryptoPunks are the creation of creative agency Larva Labs, and the sale was through UK auction house Christie’s.
🚚 Automaker VW will start testing its self-driving vans in Germany this summer. The vans are the product of a partnership with self-driving startup Argo.
🔋 One of the largest pipelines in the US was shut down after a cyber attack. The 5,500 mile pipe transports gasoline, diesel and jet fuel, and supplies the East Coast with 45% of its fuel.
🛰️ Elon Musk says SpaceX will launch a Dogecoin-funded satellite to the moon. The mission is set for Q1 2022.
🌌 Generative Adversarial Networks will allow scientists to simulate complex universes in one-thousandth of the time it currently takes. Yin Li, an astrophysicist at the Flatiron Institute in New York City, says his technique could revolutionise theoretical physics.
🌍 Humans of Earth
Key metrics to help you keep track of Project Human.
🙋 Global population: 7,865,321,403
🌊 Earths currently needed: 1.7840496708
💉 Global population vaccinated: 4.2%
🗓️ 2021 progress bar: 36% complete
📖 On this day: On 13 May 1989 large groups of students gather in China’s Tiananmen Square and begin a hunger strike.
The Language Animal
Thanks for reading this week.
Humans have always dreamed of a machine that could really talk to us. Now, via deep learning language models, we’re close.
New World Same Humans will be watching the possibilities, and the dangers, as they unfold.
And there’s one thing you can do to help with that mission: share!
Now that you’ve made it to 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 on Sunday. Until then, be well,
David.
P.S Thanks to Monique van Dusseldorp for additional research and analysis.