Reliably unreliable
Strategic innovation in times of uncertainty – a conversation with Ben Sauer
Hey there,
Do you know Ben Sauer yet? I realised I was quiet here about my previous conversation with him, published back in June. That was all about communication (or lack of) in orgs. Several people got in touch to say how much they enjoyed it. And I had such a good time that I had to book in another.
Before we jump in:
If you haven’t already grabbed it, Ben’s book Death by Screens is an awesome handbook for clearer communication.
And my new course Master Multiverse Mapping will help you figure out what you need to communicate clearly. (Perhaps you’d like a free taster.)
In the video below, Ben and I spend 90 minutes talking about AI hype, timeless innovation principles, the evolution of design research, playgrounds over goals, the management/innovation paradox, conversational interfaces, investing in AI with a barbell strategy, and more.
Our guiding questions:
What approaches and principles are sensible to adopt in times of rapid change?
What can we learn from the history of innovation to ground us?
What can we say about determinism versus non-determinism in that messy sphere that’s badly labeled “AI”?
How would we advise a company that wants to invest in GenAI?
Here’s the video (and just below it you’ll find timestamped topics as well as LOTS of links to all the rabbit holes we mention):
Tom x
Timestamped topics
00:31 Notes from UX Brighton: impossible predictions?
01:38 Ben’s talk in Romania
02:02 Imagine you’re a mobile phone manufacturer in 2006 ...
06:49 The seeds for innovation are there long before the innovation happens
09:12 Ken Stanley & the myth of the objective: stepping stones, doors and underpants gnomes
13:40 Xerox PARC, the management/innovation paradox and creating playgrounds
22:56 Running interference with deceptive visions
31:01 Fuzzy knowledge
33:45 Conversational interfaces don’t know about the world
42:51 What advice would we give to an organisation?
43:33 Infrastructure and context
45:08 You want a wise PA, you get a blundering intern
46:09 This kind of AI can’t do inductive and abductive reasoning
48:23 The reliability problem
50:32 Human oversight
54:13 Getting the right answer takes a load of hacks
56:14 So how should a company invest in AI?
01:00:30 Your optimal choice may not be available, so what options can you take advantage of? What *doesn’t* change?
01:09:46 Multiverse Mapping for pragmatic design
01:16:36 When AI works you won’t hear about it ... how can you brute force your own secrets?
01:21:30 Randomness is inefficient + Tom Chi’s rapid prototyping
01:24:08 Vaughn Tan: pairing wild innovation with perfect delivery
01:26:08 Winnowing with PR/FAQs + How Big Things Get Done
01:30:44 The machine that makes the machine
01:32:19 DARPA’s diversified barbell of bets
Links 🕳️🐇
Jef Raskin’s The Humane Interface
Ken Kocienda’s Creative Selection
Ken Stanley’s Myth of the Objective
and his ToKcast podcast appearance I referred to
Scott Berkun’s The Myths of Innovation
Steven Johnson’s Where Good Ideas Come From
Underpants Gnomes:
“The deliberate strategy is what we present to the shareholders. The emergent strategy is what we actually do.” – a Japanese CFO quoted by J P Castlin
Maggie Appleton’s Squish Meets Structure talk
Benedict Evans’ Uncanny valley of AI
AI-enabled strategy from LearnWardleyMapping
Joapen’s take on the evolutionary stage of GPTs
My Be the Algorithm card from Innovation Tactics – Front | Back
Tom Chi and rapid prototyping:
From CommonCog: only 1 in 100 ideas gets through the PR/FAQ
Bent Flyvgbjerg’s How Big Things Get Done