Antonym: The Preview Edition
All I'm thinking about is integrating AI into our work – here are the headlines.
Dear Reader
This week’s email is a relatively concise set of notes. All of my writing time, in fact most of my waking time these past few days has been taken up with Brilliant Noise’s AI-B-C jump-start workshops which we’re now running for teams. The thing about designing and running workshops is that you always learn as well as share insights. In the case of generative AI and applying it to work, everyone’s experiences are so diverse I’m learning an incredible amount. That in turn makes me want to write it all down – mainly to explain it to myself.
Also, I’m talking at an event on Tuesday and while I don’t need to prepare a big speech – I already have plenty to say – the thoughts are turning over in my mind and I’m writing hundreds of words. I think they may turn out to be a short book, or a monograph, which is so much more elegant than “e-book” or “white paper”.
There really isn’t anything as useful as the prospect of explaining your ideas in public to push you into really testing them. The following is an outline of what I’m going to be saying (though it’s 48 hours away, so goodness knows where these thoughts will have got to by then). The exact focus of the session I’m taking part in is how to get your team to start learning and using generative AI tools like ChatGPT. I first talked about this a few months ago in the Brilliant Noise webinar “AI For Leaders” (video and resources here).
Nobody knows anything. The more certain they are the less they understand what is happening.
Little picture > big picture. AI is very, very big as a topic. You need to focus. Debating when robots will rise up or become self-aware is fun but not your job.
Hesitation or reluctance is emotional, then rational. Besides the newness of generative AI, there are also a lot of negative narratives and legitimate concerns about jobs, safety. Psychological safety and careful thinking about humans more than machines is essential.
It’s not a tool; it’s a tool-maker. People will need to make their own versions or bots from generative AI systems; ones which work with their own knowledge and ways of thinking. It’s easier than you think to make these tools, but harder than you think to break the habit of expecting tech to “just work”.
It is a race. The urgency is real - feel it. The gains from working with AI will be a massive competitive advantage - into how people work is a race. There will be winners and losers. Get in the race now even if you’re slow, and get help to start making faster laps.
Performance gains are immediate and cumulative. The early evidence of advanced systems like GPT4 on productivity and performance is impressive on its own, but our experience is that the benefits accrue, at times it feels exponentially so.
How many lightbulbs does it take to change people? The importance of hands-on use and practice with these systems begins with the “aha” or “lightbulb moments” when people feel the power of thinking faster with AI.
Sharpening the saw. Structured thinking is required to get the most from generative AI. Metacognition - thinking about how we are thinking - is an underdeveloped skill for most of us but when we machines that make us think faster
“Utilise don’t rely.” This principle was shared with me by a deep-thinking data expert in a workshop. Beware: rabbit holes, sleeping at the wheel, hallucinations, bias. Critical thinking, a little scepticism even, is needed when working with systems.
Maybe by next week the above will have organised themselves into a longer article. I’ll keep you posted.
Meanwhile, if getting to the Royal Institute in London on Tuesday afternoon is a realistic prospect, tickets to the Agencyhackers “The Robots Are Coming” event are are available here.
That’s it for this week folks. Normal(ish) service will resume next week.
Antony