#DesignTheTomorrow 16/05/25 asked a panel to defend five statements, regardless of personal positions. Here are my pre-debate notes. Do you agree?
- AI-driven experiments will soon replace traditional research: False
- Adaptive, personalized interfaces should be the default standard for all digital products: True
- “Web” must evolve from static pages into immersive app-like experiences: False
- Seamless cross-platform UX: start on mobile, to web, finish in AR/VR: True
- Speed-touch-gesture-driven interfaces will eclipse screen-based Graphical UI: True
AI-driven experiments will soon replace traditional research: FALSE
Humanity is an evolving concept. If I were to ask you a moral question, you'd likely respond differently here today, in London, than you would twenty years ago. Geography compounds that variety, Western media has recently retreated to less liberal-leaning preferences than just five years ago. In contrast, AI is a ghost of opinion and understanding; a snapshot of limited context, trapped in time. All AI understanding is trained upon datasets, subtly or blatantly, resulting from moments in time.
Humans evolve individually and collectively across many axes. The same person's breakfast preferences might evolve because of societal conversation on the effects of sugars, palm oils, and processed foods. That opinion then enters the checks and balances of other smaller collectives within a family home.
AI does not have autonomy in renewing its approach. It doesn't bounce off others to adapt to a new world. It will never stop stamping its feet or hallucinating to its benefit without intervention. You wouldn't trust even the most advanced engineer from 50 years ago to troubleshoot your IT problems.
Finally, Product engineering is anchored to a capitalist market of human demands. Demands will ebb and flow to serve human interests. Research is required to meet the market where it is and how it is thinking today, and the creativity in new market creation is where humans will always excel because it is to serve their evolving needs. Evolving AI training in step with even the narrowest niche would require infinite context and expense. I don't think the ends will ever realistically justify the means.
Adaptive, personalized interfaces should be the default standard for all digital products: TRUE
Anyone who has unfortunately logged into their parents netflix account or even worse their partners, will appreciate the need for personalised interfaces.
I appreciate that Google already understands this importance in AI. A few weeks ago, my young daughter ran into another room with Gemini and could just make out the inquisitive whispers. I started to sweat as I heard her querying how Santa was able to get to every single house in the world with presents? You can imagine my relief when Gemini cheerily replied well, "That's just the magic of Christmas!"
I interpret this statement broadly past the interfaces and content of today, because context is so important. Tomorrow's personsalisations will be responsible for adapting to the contexts of a world of new users, where they are, and whatever age and cognitive ability they have (like my daughter whispering under a beanbag).
“Web” must evolve from static pages into immersive app-like experiences: FALSE
Tim Berners-Lee's first web browser was called 'WorldWideWeb', it's users were also editors. Shortly after, Nexus followed on as a passive browser. He removed interactivity because passive surfing was a different type of engagement. One channel and form of medium doesn't replace another, we need information in various forms to meet us how we require. I would not be surprised if not a single panelist could point me to their first developed 'immersive experience', as it has likely been removed from the web. Next to pictures, written text is the best form of data storage; it's survived longer than anything ever humanly conceived. I can appreciate Dickens' works as future interactive experiences, without believing that all works of literature need an upgrade to be interactive experiences.
Seamless cross-platform UX: start on mobile, to web, finish in AR/VR: TRUE
There is a delicate balance to be had between designing for today’s most practical behaviour and tomorrow’s next behaviour. Recent research into the tipping point has revealed it to be a ‘knocking point’, where you constantly knock on the door of future behaviour until eventually enough people barge through and steam on.
Preparing for this debate, I recorded thoughts in a notes app (keep), wrote on a laptop into a browser (docs), on the underground, I added some bullet points, and now, sitting here, I cite my doc using a mobile App. Great design is seamless, but no one read or wrote anything of length on a mobile phone twenty years ago. Regardless of the desire to create the journey I now enjoy today for writing, it would have been a waste of invested capital in light of the return on investment.
I have lived through future AR experiences; they are awesome and inevitable. But most of us have to build in accordance with reality.
Rather than the “what ifs,” how about the realities of ChatGPT? This morning, I read a post by a form creation company called Tally, where 25% of all traffic was coming directly from ChatGPT. More immediate innovations are already pulling us into cross-platform.
Speed-touch-gesture-driven interfaces will eclipse screen-based Graphical UI: TRUE
Speed of thought is the ambition.
- Ada Lovelace wrote theories on paper notes about conceptual machines for fun in a parlour chat with Charles Babbage, but nothing was ever made or tested in her lifetime.
- Alan Turing waited days and weeks for a calculation after his mathematics were interpreted into a physical configuration of switches and cables.
- Grace Hopper waited hours after punching holes in cards.
- Alan Kay wrote on a keyboard and moved around a mouse after a few minutes.
- John Romero designed the levels in a GUI editor while playing them!
Keyboards are a temporary crutch.
Tomorrow's interfaces will seem magical, but the most magical interface you are experiencing right now is "thought to thought". Until we get there, everything is up for grabs.
Velocity: Simulate AI participants to Instantly find basic UX problems 🔗