Weeknotes 258 - physical AI in autonomous worlds
Notions from the news on robotics, AI and connected worlds. With interesting events.
Hi, y’all!
Last week I miscalculated my vacation planning and newsletter timing and had to skip the weekly update. Returning on Monday evening and having to prepare for a workshop later that week in Hamburg occupied the calendar timewise and mentally… The workshop was very nice btw, part of the Hoodbot/Wijkbot toolkit for prototyping human-urban robot interactions. You can find a short report on the Wijkbot website.
This month, I am starting a new professional endeavour. More below at the end of this newsletter. First, let’s look at the news of last week.
Triggered thoughts
Hard Fork is definitely one of my favourite tech podcasts now. Almost every week they dive into topics that are top of my mind too and they say the right things adding the right examples, and are also very amusable to listen to.
This week's topic is about AI becoming tangible and entering physical products. Rewind, Humane AI pin, Meta Raybans. And OpenAI is teaming up with Jony Ive for the iPhone of AI. The AI promise is now mainly on the interaction making frictionless; they discuss what friction should remain friction. For me, it is more than those interactions, but also the consequences of living with an intelligence. It is something that is, of course, the foundation of the predictive relations thinking I started at TU Delft. This paper by Peicheng Guo is a nice exploration. And I feel like I need to close read this piece of Ethan Mollick: the shape of the shadow of the Thing.
Are we automating the simple stuff with stupid tools? Another angle in the post by Matt Webb this week, thinking about the consequences when intelligence is too cheap to meter. The predicted consequences of ubicomp, when almost everything we use day to day is computational, is a state we have almost reached. What if we have ubigpt? The intelligence has the same level of seamless integration. That connects to the physical AI that is starting to pop up. And to the state of predictive relations that can alienate or create another layer of consciousness. Something to explore further.
“Autonomous Worlds are not just worlds that happen to exist on-chain, but worlds that could not exist otherwise.” This event makes me curious…
Events
Speaking of events, these are some that are happening soon:
- Today and tomorrow: RSD12 in Amsterdam (”Relating Systems thinking and Design**”)**
- 17 October - Climate Tech Amsterdam
- 18+19 October - Design & AI - Shaping the Future(s)
- 18 October - Sensemakers AMS DIY
- A bit further in the future is Dutch Design Week, always a lot to do. We will be doing a workshop at DRIVE.
Notions from the News
I decided to limit it to last week; there is already so much. Unless there is some bigger story I should mention and save during my vacation… Also, I limit it to tech news, although the world has the danger of getting into a race to the bottom.
New AI capabilities: Arc browser combines OpenAI and Anthropic model, Object aware editing in Adobe AI photo, Canva Magic Studio, new Google Assistant
And mundane robots: a muffle voices. Or delegated people
In case you like to refresh your prompt engineering skills:
After finishing the writers strike, there is now some discussion on using generated AI images…
What is a photo? The computational photography on smartphones is the poster child for our synthetic reality
The success of NVIDIA stimulates other players to look into AI-focused chips. Like Microsoft and OpenAI.
Instructions for AI-assisted decision-making, how to incorporate more heuristics and mental models into your life
AI as better validator
How about scale? “The natural world shows us how to grow without leaving behind a trail of ruin.”
Choudary - who introduced an insightful model for the protocol economy - is now describing how to create value in Generative AI
Can people deal with an AI Sticker Generator?
SmartThings everywhere; hubbing your home
An autonomous wheelchair sounds like a no-brainer at first, but as a second thought, do you want to delegate that final agency?
Investing in robotics: an overview of companies
Radicle Civics aims to explore alternative approaches for organising the future, focusing on pathways encouraging distributed agency and participation from a super-diverse public to challenge concentrations of power and responsibility.
CJ Trowbridge is doing an interesting analysis on TikTok. On Musk mainly, but also more generic. Or sometimes more focused on philosophy in general…
On the list for my next vacation:
Or maybe the Sphere? Kelly is describing on the platform aspects of this attention-grabbing object.
Paper for this week
Beyond data transactions: a framework for meaningfully informed data donation
This model for literacy in data donations could go beyond use by researchers and designers.
In this paper, we investigate how to foster (personal) data literacy and increase donors’ understanding of their data. We introduce a Research through Design approach where we define a data donation journey in the context of speech records, data collected by Google Assistant.
Gomez Ortega, A., Bourgeois, J., Hutiri, W.T. et al. Beyond data transactions: a framework for meaningfully informed data donation. AI & Soc (2023). https://doi.org/10.1007/s00146-023-01755-5
See you all!
As mentioned above, my daily activities will shift. I have been working on shaping the design principles of the AI co-pilot, among others, for Structural in the last year. New funding is needed to continue the development, and in the meantime, the company is in hibernation. I will use the opportunity to focus on Cities of Things and Wijkbot activities (and not to forget ThingsCon), and extend future research activities and product discovery via a new company with an existing name: Target_is_New. More on that later in detail, but find some backgrounds via the targetisnew.com. And I love to have a chat to tell you more!
See you next week!