Fix your broken UI before you build your shiny AI
I attended a couple of conferences over the last week-and-a-bit. They were good. Good content, good speakers, good side conversations, good food. I even spoke at one of them. Super.
They were both very different types of event but it was intriguing to see similarities in how they pitched their wares. Almost every session I attended ended up being about – or veering into – Artificial Intelligence, which meant speakers were compelled to shoehorn a reference to AI into their slides or their chat (me included).
As a consequence, it could be tricky to work out if presentations were referring to automation, machine learning, LLMs, AGI, robot obliteration of the human race, self-generating cat videos, or a mix of all of the above – a point Scottish Government’s Permanent Secretary JP Marks observed as he closed Digital Scotland 2023 (not the cat videos bit).
Despite the AI focus being somewhat overwhelming, the content on the whole was helpful, nuanced, and thought-provoking. I was struck by the open discussion of ethics, privacy, checks & balances, the need to approach any long-term solutions with caution, and the importance of finding the correct use cases.
There was some nice emphasis on getting the foundations right, and encouragement to spend time experimenting/using a co-pilot before leaping at full pelt onto some vague transformative solution. Y’know, get to know how something works before banking your organisation’s future on it.
As I’ve written before, my gripes about AI aren’t really technology based – it’s the hyperbole surrounding it and the difficulty of detaching from it where I see problems: We all keep on talking about the thing even though we want to ignore the thing but the thing keeps on tapping us on the shoulder and is like hello I’m the thing and oh no it’s almost as if the thing is ghost writing what you’re reading RIGHT NOW…
Last week, one of the conference corridor chats I had involved a person lamenting the relentless push on a product’s new AI features, which they felt were coming at the expense of the user interface (UI) and overall user experience. The way they spoke really resonated with me, and linked in with some of the stuff that’s spinning round my little brain. And 🎉 ta dah! 🎉 the title of this post was born.
Because AI currently feels like the collective hook we can’t avoid hanging our coats on, I purposefully want to spend some time proposing some areas of focus that are definitively not AI-based (but conversely might be ways in which your organisation becomes AI-ready, should that be your route of choice).
I guess my intent here is to lightly prod the sectors I hold close to my heart – the public, third, and culture – where I can see in real time the AI conversations taking hold, and where I’m gently suggesting there may be matters more pressing.
With money and resources getting tighter and tighter, the temptation to focus on shiny miracle solutions is probably greater than ever. But shiny miracle solutions do not a strategy make and addressing endemic issues, while much less exciting, is where some of that energy would be better spent. Fix your plumbing and unleash your future growth potential (or something). At the very least you’ll keep Mario happy.
I should clarify, when I’m using ‘you’ or ‘your’ in my writing, I don’t mean 🫵 you – I mean me, I mean we, I mean us. There’s a collective responsibility here to subtly shift emphasis, challenge some of the narratives, and move conversations in different directions. Technology is always a work-in-progress and there’s never going to be a point it all falls neatly into place. The important thing is to get the fundamental direction of travel right, and have the ways and means of staying on track or re-orientating if you need to.
I’m going to spend the next few weeks putting some thoughts down about where I feel some of that AI verve could be alternatively invested by organsations. My working list is here, but I may nip and tuck it as I go. And as ever I’m always open to thoughts, suggestions or challenge.
Why your online recruitment ‘portal’ should be top of your list of things to fix
It’s probably time to revisit your social media strategy
Beef up your cybersecurity
Get your data and processes in order
Stop reading long opinionated posts about AI, like this one
Questions of ethics
To kick things off, I thought I’d link to a few useful resources about the ethics of AI. Unlike the more practical topics directly above, I would definitely advise spending some time getting a more philosophical understanding of where some of the real issues are (rather than, say, the OpenAI psychodrama that’s played out over the last week).
I’ve found the following pieces fascinating and provocative. I don’t necessarily agree with all of the standpoints, but they’ve helped me develop my own understanding of issues, and the questions it’s important to ask upfront.
Has technological innovation lost the plot? an interview with Dr Shannon Vallor;
Ezra Klein’s interview with Kelsey Piper: AI is about to get much weirder, here’s what to watch for;
Rachel Coldicutt’s post On Understanding Power and Technology;
AI in healthcare, hope or hype? podcast with Sir John Bell and Dr Axel Heitmueller;
Why we must resist AI podcast with Paris Marx interviewing Dan McQuillan;
Resist the urge to be impressed by Emily Bender.
📖 Thank you for reading.