Weeknotes 318 - Dreaming about regenerative AI agents
Last week’s ThingsCon brought lots of inspiration, among others in the talks by Matt, Iohanna, and Simone. Some first thoughts. And the tech giants battling for the hottest news.
Hi, y’all!
Indeed, last week, I did not send a newsletter. It turned out to be too busy to pay enough attention to the newsletter. The organizing of ThingsCon’s TH/NGS 2024 conference was the reason. The special edition for the 10-year celebration, including a dedicated exhibition on Generative Things, took some final sprints…
It was a great edition, a reunion of the ThingsCon community, and especially an inspirational moment with all the work of students, fine workshops, and talks. And a great party. Thanks Peter, Iohanna, Alex, Joseph, Peet, Matt, Oana, Alina, Kai, Henrik, Sen, and all the others for sharing their experiences in person, for the kind words!
Pictures and videos will be shared as soon as possible, and I will take some time in the holidays to write some reflections on the experiences. We plan to follow up on the theme next year with the exhibition and a new edition of the RIOT publication. Let me use some cues from the three talks on Friday for a triggered thought.
Triggered thought
Where are we now? Simone Rebaudengo took us on a trip through ten years of technology development. Agents, agents, agents. Just like Matt Webb earlier, he stresses the importance of connected agents. Designing is now the design of the relationship and organization between the agents. Not the personality but the system where the agents live.
He refers to Venkatesh Rao: If you build an economy inside things, your control will go from deterministic to stochastic, and maybe it is overkill, but you’ll create true creative complexity. (Venkatesh Rao, 2024)
With the Walkcast, he is already experimenting with how a team of agents can become a micro-studio for creative work. He ends with his beliefs for the next things we will make:
It’s the time for small, custom, niche, crafted intelligent things. It’s the time to build non-linear, spatial, unclear, post-conversational tools. It’s the time for complex, surprising & non-deterministic experiences.
It’s the time to design worlds instead of software inside things.
Matt Webb also used the world as a metaphor for the thinking in personal context windows that drive the new things we will develop. A world of agency, where we get self-driving mode in every system. Becoming an adaptive world of mini-apps that make easy things become trivial. In a world of intelligence too cheap to a meter, it will bring ChatGPT-like intelligence into everything, creating a world of tiny personalities with AI affordances. He is ending with a world we must dream. We are aiming to get the strangeness out of this intelligence while this is delivering opportunities to open up new worlds. We miss what this generation of intelligence makes special. Dreaming, invention, stories, all kinds of hallucination. New things happen on the other side of the bridge.
Are there links to how Iohanna Nicenboim aims for a balance between human-centered AI and more-than-human design to reach a more sustainable and inclusive future? Something we might call Regenerative AI, creating conditions where both humans and non-humans can thrive. Twin intelligence. Technology shapes what it means to be human, and also what it is not.
With all the attention given now to agentic AI, and all the expectations for this as the dominant trend for the coming year(s), understanding that we are designing the world of agents in things, that can inspire new paths while we stay aware of the connections we make, steering it and letting it steers us. Not towards a certain new tool for general intelligence, but for more humanity.
For the subscribers or first-time readers (welcome!), thanks for joining! A short general intro: I am Iskander Smit, educated as an industrial design engineer, and have worked in digital technology all my life, with a particular interest in digital-physical interactions and a focus on human-tech intelligence co-performance. I like to (critically) explore the near future in the context of cities of things. And organising ThingsCon. I call Target_is_New my practice for making sense of unpredictable futures in human-AI partnerships. That is the lens I use to capture interesting news and share a paper every week.
Notions from the news
These past weeks have been very heavy on new product introductions, both by OpenAI in a (still running) 12 days news advent calendar and Google introducing all new Gemini features.
In the meantime, in the US TikTok might be banned, or will it become a political football (is that the right translation?)
Human-AI partnerships
Agents, agents, agents. Indeed. Google Project Mariner “A research prototype exploring the future of human-agent interaction, starting with your browser”. And Google Agentspace for ai-powered search. And research. And Android XR is for Gemini-infused wearables. And more mundane features.
Mapping out in this AI Agent Market Map.
OpenAI officially launched Sora (not accessible everywhere in the world though). And ChatGPT search.
Meta is rolling out live AI to their Glasses.
Anthropic has a tool for moderating it’s chatbot, called Clio.
Robotic performances
GM is dropping out Cruise self-driving taxis. While Waymo is sending their vehicles into Japan. And China is continuing to grow its fleets.
“As robots become more complex and natural in their movements, our understanding of what it means to be human is also evolving.” We are the robots.
Immersive connectedness
The new social nostalgic by Ev Williams is called Mozi. Someone refers Dopplr. Or jumping in where Foursquare jumps out?
Also from the Gemini news: Nest refers.
Tech societies
Klarna is changing it’s hiring policy without showing, thanks to AI.
Ed Zitron is opposing the digital world harming users because of companies prioritize growth over user experience.
Sacasas calls for greater self-awareness and humility among tech developers to ensure their work genuinely empowers and serves communities.
Will the EU follow fast enough?
2025 ahead
It is that time of year: gadget gifts.
And looking back at the gadgets of 2024.
And predictions. Azeem Azhar with Ethan Mollick, looking ahead to 2025. From individual tools to team tools. AI will drive companies into a learning mode again. Will it be secret or open-source?
Ilya Sutskever predicts new ways AI will be built.
The Verge is having a podcast series on 2025.
Foldable iPads, will they be real?
Paper for the week
On a method. Using Annotated Portfolios to Interrogate Speculative Designs: The Case of Emergent Personal Data Trails
This paper is concerned with speculating new and emergent personal data trails, and how an Annotated Portfolio approach might help us understand how to support users’ in this space.
Accordingly, in work anticipating and speculating about new data interactions or emergent sources of personal data, we need to closely consider their potential for mosaic analysis. The Annotated Portfolios method offers great potential in this regard, representing a means of exploring the relationality between designs to gather new knowledge about the portfolio as a whole.
Snow, S., Khan, A. H., Matthews, B., Rodríguez, I., Desjardines, A., Seals, A., ... & Viller, S. Using Annotated Portfolios to Interrogate Speculative Designs: The Case of Emergent Personal Data Trails.
Looking forward
This week is dedicated to completing a report for the research project.
There are not that many events this week I guess, but you can attend a Sensemakers XMas special. Building your sensing xmas gifts.
Next week I might skip for holidays, or have a short edition depending the news.
Enjoy your week!