Some things at the intersection of technology and society that will probably happen in 2023
Earlier this week, the 2012 Horse ebooks tweet “Everything happens so much” popped up again in my timeline, like a 10-year anniversary special. And it’s true: everything’s been happening so much lately, I wanted to pull together some things I’ve been thinking about and poke a bit at the overlaps and commonalities between them.
If there is a theme that runs through the following, it’s that 2023 will be full of quite mundane-seeming change. While we’re waiting for the next moonshot to save the planet or the next billionaire to go bust, we won’t much notice that our chicken wings are being delivered by a robot, but ultimately this kind of tech normalisation will have a bigger impact on the day-to-day reality of most people’s lives than any of the things peaking out on the hype cycle. I’ve grouped the observations under four headings — People working with machines; Polarisation and decentralisation; Inequality; and Regulatory entanglements — but there is plenty of overlap between those topics.
Also, a caveat: I’ve been ill for most of the last month (which, let’s be honest, feels very on trend) and am left with a bunch of lingering Covid symptoms that make concentrating hard, so I started this as a writing exercise and am now (in another very on trend, post-Twitter move) sharing it as a blog post. I’ve not run any of these thoughts through a fancy superforecasting process, it is just a note of how the world appears to me right now.
1. People working with machines
The generative tchotchkes will go mainstream
2022 has shown us that a lot of people really like playing with generative tchotchkes like Dall-E, Lensa, and ChatGPT. There’s something thrilling about the low-stakes death drive experimentation made possible by these limited attempts at human replacement, and they’re also a great vehicle for making clever jokes on social media.
Meanwhile, any technology that can render evermore infantilised, sexualised and suppliant depictions of women is always going to generate huge amounts of media coverage in which everyone looks extremely shocked about how on earth those boobs got so big on the made-up programmable lady servant.
Shipping these kinds of tools is a win-win for any company struggling to find market fit as an AI/metaverse start-up: they generate huge amounts of PR and name recognition while also delivering proof of buzz to investors. As such, it seems probable we’ll see a lot more of these tchotchkes in 2023, and they’ll rapidly settle into the free and very cheap cloud-based kit of parts that companies use to deliver their digital customer services.
Robust quality assurance models for these kinds of tools are still a long way off, and large data and language sets tend to normalise and encode the world view of a very small subset of society,* so it seems sadly probable that 2023 will see more than its fair share of Bots Gone Rogue. If I were a betting person, I’d put money on seeing some very stark racist, ableist and sexist outcomes from the overdeployment of these technologies very early in Q1 next year.
*There is plenty of accessible scholarship on the topic of bias in large data sets; Vinay Uday Prabhu and Abeba Birhane’s “Large image datasets: A pyrrhic win for computer vision?” is an excellent recent starting point.
Human customer service will start to denote artisanal quality or high-end luxury
Since the start of the pandemic, fast casual dining chains have been hiding in plain sight as pioneers of (often low-tech) automation. It’s probably only a matter of time before your Nando’s order arrives at the table on the back of a delivery robot, and this year UK chains including Leon and Itsu have been trying (not that successfully) to edge towards staffless, Amazon Fresh-on-a-budget self-service customer flows.
Just as in small-format supermarkets, where staff duties include servicing the needs of the grocery delivery apps and resolving the failures of the self-service checkouts, service staff in chain restaurants are likely to become machine monitors: there to manage fail overs and glitches rather than to cultivate relationships with customers.
Two outcomes of this will include the continued rise of the phenomenally boring job, in which humans are required to provide continuous partial attention to apps and robots, and the emergence of a very weird, slightly edgy ambience in low-staffed, unsupervised chain restaurants. Meanwhile, human customer service will come at a premium — a sign of either a small business struggling to get Zettle to work for cashless payments, or of a high-end luxury experience.
2. Decentralisation and polarisation
VCs will keep pumping money into Web3, missing the whole point of decentralisation
The Careful Industries team conducted a piece of research for the Frontier Tech Hub this year on how Web3 might contribute to solving the great global challenges. The closest we got to an optimistic answer was that — like many other new technologies — Web3 would deliver an “innovation uplift”, bringing new money and energy into the service of complex and existential problems.
It’s possible we disappointed our commissioners in failing to find a silver bullet application for these technologies, but we did spot something else: the fact that people are tiring of relentless centralisation, and are looking for new ways to gather and organise. In some cases we noted this as a response to the real and perceived failure of existing technical, political, and social infrastructures; in others, it seems motivated by a desire to find new ways of distributing power and assets in the face of climate injustice, deepening poverty, and financial exclusion. These social shifts seemed more significant to us than any of the technical attributes of Web3, and we explored the potential long-range social and political implications of this in our report.
Whatever the case, this urge to regroup is fundamentally at odds with the tendencies of investors to massively overinflate tech markets in search of scale. Overvalued tech bubbles are highly likely to continue to pop next year, but the most significant social shifts to decentralisation will happen well beyond the bounds of start-up culture.
We’re going to get more niche
Three decades of the World Wide Web have made it much easier for large numbers of people to find affinity with people who share their hobbies and interests.
As life gets harder and more expensive, finding deep affinity will become much more appealing for some, and we’re likely to see many more niche interest groups come to public attention in 2023. As trend forecaster Sean Monahan put it in The Guardian this week:
In retrospect, I believe the vibe shift is a return to fragmentation. Culturally, the 2010s were an era of centralisation. Subcultures died. Instagram reigned supreme.
The proliferation of niches and subcultures is different to polarisation, and not just restricted to fashion. The Culture Wars-ification of political discourse means that widespread polarisation has become accepted as a fact of post-2016 life, but the term is often misused to refer to any issue on which there is a lack of clear public consensus. However, failure to reach agreement is not necessarily an indicator of polarisation; it might also be a sign of diversification, reflecting a plurality of opinions or preferences that cannot be reduced to a headline or an answer in an opinion poll.
Defaulting to polarisation also assumes that all human life is arranged on a linear, left-right scale, but traditional party politics is only one (fairly limited) way to understand the world. After all, much of human life takes place outside of the artificial bounds of political affiliation.
There are, of course, a few hot button issues that provoke fierce binary opinions, but these tend to be outliers rather than indicative of the general state of popular opinion. As Bobby Duffy noted at the end of last year:
people of all generations are identifying more with their own group and differentiating themselves from the ‘other’ group
This process of identification is not always antagonistic, and overall it seems probable that we are actually shifting towards a more decentralised society, in which there are many different groups, all of which have greater affinity with their near neighbours and perceive significant differences with others.
Many factors will have led to this — including three years of pandemic life, an increased reliance on algorithmic content curation, and broader generational shifts in media consumption — but it is likely not an entirely new phenomenon, but the visible endpoint of a slow shift over decades.
One potential, optimistic outcome is that increased decentralisation could eventually drive a more inclusive society in which more people are free to make choices, more perspectives are deemed to be valid, and fewer people are expected to conform to the norms of the dominant group. Adapting to this will be difficult for traditional organisations and perhaps violent at some points, but the end result may well be that more people have the means to flourish and be self-determined. There are many challenges in the short term, and it’s likely 2023 will see significant cultural friction in response to increased visibility for more niche and subcultural groups.
The billionaires will be forgiven, and get a second and third chance to spoil it all
Maybe not Sam Bankman-Fried*, but unless Elon Musk loses all of his money or goes to Mars, someone will keep handing him free passes.
*I mean, who am I kidding? Probably also Sam Bankman-Fried
The Hostile Environment will continue to be a testbed for inhuman uses of innovation
There’s not much more to say about this, and I don’t want to give anyone ideas, so I’m going to leave it at that. However, in case you can’t imagine what could be worse, this Mastodon post from Damian Williams and this from Evan Greer are two examples from outside the UK of ways “innovative” infrastructure can be easily misused.
4. Regulatory entanglements
Lastly, as 2022 comes to an end, there is unfinished digital regulation legislation on the books in the US, Europe, and the UK. How these different regimes interoperate will start to become clearer next year, and some of that will get pretty messy. This will have all kinds of consequences, but one specific thing is that it is likely to be tough on the digital rights sector: I hope I’m wrong, but during this period of adjustment it seems possible it will be even harder to get work funded that seeds alternative hopeful futures, as more effort is funnelled into supporting or responding to the legislative shifts. Of course, both things are important, but it’s just as important to start new things as it is to stop existing ones.