The new year is a week old now. High time for a traditional look into the year to come. Last year we came out of COVID lockdowns, or were even still in the last legs of it at this time of the year. That impacted our lives also in relation to the surge of tech in our daily life for everything. And then at the end of February, the world was shaken up by a war on this continent (Europe). It impacted the positive mindset and economics; I am not telling anything new. Weirdly enough, the war did not go fully digital but was back to old-fashioned ground war. The intelligence and the communication technology is more key than ever though. In that sense, it is a hybrid war.
I do not want the predictions built around this though. I am afraid it is more likely it will take longer than shorter. Zooming out, there is also some longer shift happening towards a different focus in the world order, especially with China slowly but steadily expanding interests and influence. I doubt we will see any big shifts this year, but it is definitely something to keep in mind. One of the practical impacts is our attitude to data-driven devices and services. TikTok is the most known and highly discussed if regulations are necessary. It requires new thinking for sure, and it is not limited to our social media; also the new wave of cars that have much more sensors create an environment that is listening more than ever. The same is true for all kinds of consumer goods and infrastructures.
TikTok is a poster child here. Just last week, there was a case of good old espionage, but what is much more pregnant is the building of an influence machine. I don’t think the core problem is having Chinese collecting data on people; it is all about the possibility to control the algorithms that make TikTok as addictive as it is. In peaceful times, it sugarcoats your view on reality in a kind of positive way; it is also shaping a reality that is not there. It shapes new bubbles. It is easy to interfere in these bubbles. One of my first public presentations at Reboot back in 2008 I was exploring the thoughts of virtual gated communities. Algorithmic bubbles are the new form, only the guards of the boundaries are not the people but the rulers of the algorithms…
On the positive side, we see a backlash that is triggered or accelerated by the acquisition of Twitter by Elon Musk. The federated model of Mastodon is becoming much more a serious alternative for the real social part of social networks. I don’t think that Twitter will disappear, even if it goes out of business, it will be bought and rebuilt, I expect. The nature of the networks run by big tech have been moving towards publishing (or shouting) instead of building communities. We are looking for alternatives for the communities in tools like Mastodon that has a first mover advantage, and also still in closed communities like WhatsApp groups.
That is a positive development that might become even more important in 2023. I am curious to see if this also trickles down in a new spring for Web3 and DAOs. I expect we might become more used to the background principles for building the federated, common-based communities and if it grows in importance, the autonomous part will become also key. I wrote a research paper on the topic at the end of last year and the idea of Decentralized Society (DeSoc) is something to keep track on. Being realistic, it will in 2023 be limited to some experiments mainly.
In that research, the framing was not on Web3 so much as it was on the protocol economy. It is a nice frame for what is happening on multiple levels. In design, it means we need more focus on the design of rules and protocols; that is becoming more defining for the experience of using products and services. The design of the touchpoints remains important, but the groundwork in the protocols is defining. That is of course the case with algorithmic-driven experiences but also in general in the more complex society we are designing systems and infrastructures for. For me personally, this connects a lot to both the Cities of Things as my new main focus, working on the design language STRCTRL.
One example that just becomes very relevant in 2023 is the uptake of Matter as a new standard for connected things in the smart home. The different big players formulated a standard protocol that is creating a new playing field, a design canvas for new products. At this year’s CES, it was the main new development (next to adaptive colouring cars :) ).
I hope we will not only see big tech dominating this standard, but it can trigger new ideas like the app stores once did. Having the right security and privacy that not only set rules on data but also on the applying of the protocol in algorithmic behaviour is needed. I expect this will be a hot topic at this year’s ThingsCon events (but that is a prediction easy made…).
Speaking of algorithmic behaviour, there is no 2023 outlook possible without touching upon the biggest shift of last year: GPT-3. Started with the imaging tools of DallE2 and Midjourney creating enthusiasm, with the opening of ChatGPT to all users creating instant excitement, and also triggering a lot of questions about what it will mean for 2023 and beyond in our relation with AI.
Especially if you have been focusing on the relation of machine and human intelligence in working together to solve problems and if you have been following the developments of AI and robotics for some years. The GPT-3 model was initially released in 2020 but with the introduction of a chat interface and the richness of that dialog, it really got the impact. Only since November 30, 2022, and it created a huge amount of applications and ideas. ChatGPT has taught us that the dialog is more important than the knowledge, but mostly that the dialog is part of the intelligence. Co-performance is the future.
We are just at the beginning. We see some things work better than other things, and it is limited in the base for the knowledge. With extending that to real-time and the introduction of GPT-4 this year, it will remain in the center of attention in 2023. Google will probably also introduce (parts of) its own system LaMDA. Back in June, the general response to the story of an engineer thinking the systems came alive was still skeptical; I think most have changed that opinion now.
Some of the questions for 2023 for these developments I have:
- Will we see a connection with real-world devices and experiences? In two directions: back to Matter: the connection of things to the internet will become more mundane than before. How will the AI dialog be integrated and define our experiences?
- And the other way around, we see a slowly increasing use of AR/VR devices for specific use (gaming) and if Apple will introduce its AR platform with WWDC in June and a device some other moment later this year, how will that make use of alternating and enhancing the reality we experience and create a new type of AI bubbles? In 2024 at the earliest.
- What are new domains for protocols to design? Especially on the city level, we are discussing these for times, but creating new types of neighborhood communities as an overlay might become more reality after the first specific experiments.
- How will prompt design and interfaces that prompt us develop into the leading design discipline?
- Will we have more AI enthusiasm, overtrust, or AI scepticism after a year of experiments and new product introductions?
On the latter, I think this year will be a defining year for our relationship with intelligence technology. We have discussed a lot on these topics in the last few years, on how we should strive for co-performance more than seeing tech as a tool. It is worth revisiting that research. I also think that the discussion (and research) about contestable AI by design is more timely than ever. We can even go back to the theories of Engelbart on augmenting human intellect… And in my own explorations on changing perspectives when we deal with predictive intelligence and things that predict.
The thinking is not new, but experience for real makes a difference. In 2023, we will see how opening up these technologies will lead to a lot of new learnings on our relationship with technology, but also will bring both small and big accidents.
Our biggest challenge might be the further increase of a synthetic reality we will live in, the algorithmic bubbles that blur our view on what matters and what we should defend. We claim the importance of authenticity, but we are rapidly building a filter between reality and experience, in our phone cameras, in cars, and in dialogues with machines. I wish for 2023 to be a learning year that prompts our thinking, and keeps the scenarios open to apply the learnings in designing our preferable future…