New Week #115
Alphabet builds a next-level robot brain. Denmark has an epic plan to ease global heating.
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 24,000+ curious souls on a journey to build a better future 🚀🔮
A shorter instalment this week. Here are three quick stories that blipped on to my radar; each of them circles around a longstanding NWSH obsession.
Alphabet are questing after an all-purpose household helper robot. Denmark is launching an epic project in geoengineering. And generative AI is colliding with the global economy.
🤖 Bodies and brains
Back in New Week #95 last summer I wrote on the amazing work happening inside Alphabet’s Everyday Robots division. This week, a development.
Researchers at Everyday Robots announced PaLM-E, which they call an embodied multimodal language model.
Essentially, they’ve taught Google’s massive PaLM language model to ingest ‘raw streams’ of visual data from sensors placed on the body of a robot. The resulting ‘general-purpose visual-language model’ is able to understand natural language instructions, and navigate its way around an ordinary, unstructured physical environment.
In short, you can tell this robot to ‘bring me rice chips from the drawer’, and it will understand you, hatch a plan, and execute.
⚡ NWSH Take: The first-order implications are clear: step by step we’re getting closer to the kind of all-purpose, household helper robot that we’ve all dreamed about while loading the dishwasher. For more on Alphabet’s work towards this goal, go back to New Week #95. // But there are deep philosophical implications, too. The endgame for AI research has always been an AGI: an ultra-flexible, human-like artificial intelligence. Critics point out, though, that it’s impossible to separate human intelligence — and certainly human consciousness — from our embodiment. From our experience, that is, as entities with bodies in constant contact with our physical environment. PaLM-E is an attempt to bridge that gap — to create an artificial intelligence with its own body. At some point we’ll have to ask: what is the difference between an embodied AI and an animal? For a deep read on that and related questions, see this essay.
🌊 Bring us your carbon
This week, news that Denmark will become the world’s first country to import and store CO2 captured in other nations.
The government-sponsored Greensand project will see the creation of a CO2 graveyard 1,800 metres beneath the North Sea. Engineers hope the reserve will store up to eight million tonnes of CO2 per year by 2030; that’s more than 13% of Denmark's annual CO2 emissions.
The project is the work of a consortium of over 23 Danish and international partners, including INEOS Energy.
⚡ NWSH Take: Yet another signal, then, that technological solutions to global heating are now firmly within the mainstream. Last week I wrote on a new UN report that says scientists should find safe and reliable ways to reflect the sun’s rays away from the Earth. And this week, news that a team of US scientists is doing just that: they’ve deployed a modified B-57 bomber over the Arctic to measure trace gases in the atmosphere that reflect sunlight. The aim to is to develop a better understanding of how those gases could be used to cool the Earth.
💸 Cash machines
We all know that right now generative AI has main character energy. This week delivered multiple glimpses of the transformative impact these AI models will have on the economy.
Microsoft continued its race to integrate ChatGPT with its applications — software that fuels working life for hundreds of millions of knowledge workers globally.
It unveiled Dynamics 365 Copilot, a ChatGPT-fuelled tool for workers in sales, marketing, and customer support that assists with meeting summarisation, drafting emails, content creation, and more. The tech giant also announced that the new AI-juiced Bing has passed 100 million daily active users; a major milestone, albeit still a long way from Google’s 1 billion DAU.
And speaking at an event today (Friday), the Chief Technology Officer at Microsoft Germany, Andreas Braun, said the company will release GPT-4 ‘next week’. Braun told his audience that GPT-4 will be multi-modal: able to respond to text prompts with text, image, video, or music. At the time of writing, Microsoft have not confirmed this statement; let’s see what next week brings.
Meanwhile, Character.ai, a startup that leverages generative AI to allow users to build and chat with their own virtual companion, raised investment at a valuation of $1 billion. The $200 million round was led by Andreessen Horowitz.
Character.ai was founded by Noam Shazeer and Daniel De Freitas, two former Google engineers who helped build the company’s LaMDA conversational AI.
And the iconic Citadel hedge fund announced that it is in negotiations with OpenAI for a company-wide ChatGPT license. Citadel became the most successful hedge fund in history after making a record $16 billion in 2022.
⚡ NWSH Take: Two recent studies add context to what we’re seeing unfold right now. In February, an MIT study on the impact of GitHub Copilot on coders found that it increased productivity by 55.8%. And in March, a study of the impact of ChatGPT on ‘mid-level professional writing tasks’ — such as writing memos and strategy documents — found that professionals who used the tool completed tasks on average 37% faster, and with improved average writing quality. These are revolutionary productivity leaps. If they are replicated in real offices then knowledge work is about to be transformed.
Thanks for reading this week.
The dream that is an all-purpose household helper robot has inspired science-fiction writers for decades. Now, it’s edging closer.
This newsletter will keep watching. 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,
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.
It's becoming clear that with all the brain and consciousness theories out there, the proof will be in the pudding. By this I mean, can any particular theory be used to create a human adult level conscious machine. My bet is on the late Gerald Edelman's Extended Theory of Neuronal Group Selection. The lead group in robotics based on this theory is the Neurorobotics Lab at UC at Irvine. Dr. Edelman distinguished between primary consciousness, which came first in evolution, and that humans share with other conscious animals, and higher order consciousness, which came to only humans with the acquisition of language. A machine with primary consciousness will probably have to come first.
What I find special about the TNGS is the Darwin series of automata created at the Neurosciences Institute by Dr. Edelman and his colleagues in the 1990's and 2000's. These machines perform in the real world, not in a restricted simulated world, and display convincing physical behavior indicative of higher psychological functions necessary for consciousness, such as perceptual categorization, memory, and learning. They are based on realistic models of the parts of the biological brain that the theory claims subserve these functions. The extended TNGS allows for the emergence of consciousness based only on further evolutionary development of the brain areas responsible for these functions, in a parsimonious way. No other research I've encountered is anywhere near as convincing.
I post because on almost every video and article about the brain and consciousness that I encounter, the attitude seems to be that we still know next to nothing about how the brain and consciousness work; that there's lots of data but no unifying theory. I believe the extended TNGS is that theory. My motivation is to keep that theory in front of the public. And obviously, I consider it the route to a truly conscious machine, primary and higher-order.
My advice to people who want to create a conscious machine is to seriously ground themselves in the extended TNGS and the Darwin automata first, and proceed from there, by applying to Jeff Krichmar's lab at UC Irvine, possibly. Dr. Edelman's roadmap to a conscious machine is at https://arxiv.org/abs/2105.10461
Thanks for this post! Much appreciated.