Hot from the oven: notes on Day One of the LSP International Conference

…and when I say they’re hot from the oven, I really mean it: I’m currently sitting at the bar of the Legoland Hotel in Billund, freshly changed and waiting for dinner (I know, it’s barely six, but it’s Denmark after all). I’ll probably do something more structured, but I wanted to share with you some notes on this amazing first day while I still have thoughts fresh in my mind and bricks hot under my fingertips.

So, this is my fourth time attending this kind of conference and what I love about it – aside from the obvious networking and a chance of being a participant in workshops, instead of a facilitator – is the intellectual level of the debate that’s set up by Inthrface and, this year particularly, by Johan Roos. My favourite so far has been the 2021 edition on organisational8mprovisation (you can find some thoughts over here), but this has every potential to top that off. Why?

Human Magic

The conference’s theme is organised around Johan Roos’ latest book Human Magic: Leading with Wisdom in the Age of AI, and that’s how we kicked off in style, after a brief introduction by Keblovszki from the Legoland Hotel, who shared how he’s constantly seeing children using LEGO to build ideas, create understanding and ask better questions.

Johan delivering the Keynote wih his characteristic, wide-eyed enthusiasm and genuine marvel at the splendors of the human brain, of the power of play and – why not – at the transformative age we live in. I want to be Johan when I grow up. And his question is: how do we design and deliver experiences in the world of algorithms? Hold on to your “human magic” is the answer.

He framed himself as a believer in the revival of books and talking to each other, before diving right into the conference’s main topic: the impact of AI on our human capabilities, and how LSP (and the power of play in general) can be a powerful mean to use these tools so that they amplify our abilities to strategise, to ask better questions, to do exactly what Keblovszki said children manage to do with LEGO.

Traditional strategy is historically stale: we arrive at the same things year after year. Artificial intelligence amplifies this, as usual. A complete, complex strategy can be now made in under 3 minutes using an LLM, and those products are going towards sameness, towards the median of AI training data. Individually we can become super creative thanks to these tools, but collectively we’ve become less creative. How is that possible? If you think about it, eery time we ask a question to an LLM we’re not asking that same question to a colleague (out of a need to save time or out of some other, less noble impulse), and as a collective we become more stupid.

That’s a fact you can’t argue with.

Queue the history of LSP since 1997. The assumption when Johan and Bart stepped into the room, back then, was that LEGO executives were super-creative… and yet they were as boring as anybody. Until the materials were brought into the centre of the discussion instead of flipcharts and traditional means of discussing strategy.

Human Magic, therefore, is about the micro-decisions we make every day in using AI tools. It’s either we become more curious, through these tools, or we stick to the first answer we get. As such, our toolbox is made of three categoripy of things: those aimed at preventing, thise aimed at mitigating, and thise we can use to amplify. Preventing means keeping the capability in use to prevent erosion, mitigating means slowing an erosion that’s already under way, and amplifying means use AI to boost your capability.

As it happens, 3 out of 4 people in big corporations are delivering AI-generated nonsense in official reports because they stopped to think critically. In the words of Johan himself, “Erosion [of human capability] is soothing, invisible, and can be devastating.”

These tools also change how we communicate: you might find yourself to use it to communicate in a way that makes you sound like everybody else.

In light of this issues, human capability of making good decisions is about practical wisdom, which is judgement in action, and I think it’s a brilliant way to define it. Here’s four critical judgments not to outsource:

1. What matters?
2. What’s right?
3. What works?
4. What resonates with my team?

The impact of AI is such that:

• we follow it also when it’s wrong because we trust it;
• the first rung of the social ladder is thinning: you put tools of experts in the hands of novices without the tools to become an expert;
• we delegate more and communicate less.

Again in the words of Johan, “Judgment is formed by doing the work. That is the part we are handing over.”

So, how do we keep our skills sharp? As we know, curiosity is one of the pillars, and we can identify two kinds of curiosity: Laser curiosity vs. Radar curiosity., where kaser curiosit is going really deep into something while radar curiosity is casting your net broader to catch many aspects of reality. And both are awesome, and the joy in exercising each is a fundamental part. What you need to be mindful of, however, is that there are cultural differences, even at a nuclear family level: some people (some genders also, I’d say) are rewarded and encouraged to be curious, while others aren’t. When a boy is curious, you usually reward it. When a girl is curios, you tell her she’s being a monkey. Tell me I’m wrong. Fortunately, curiosity isn’t a personal quirk: it’s a professional capability and it can be trained by “making the familiar strange”.

When it comes to using AI in LEGO Serious Play, Johan doesn’t assume: hi insight comes, scientifically appriached as usual, from a well-documented trial run of workshops, a group using chatGPT as much as they wanted and another group not using it. As it tuened out, people using it reported split attention when used in the room, but a huge help before and after the workshop.

Bottom line is: prepare with AI once you have a clear framing of your challenges, perform with presence, follow up with AI. Logos can be shared: ethos and pathos can’t.

Since 2016 this is what the research shows: procedures travel easy; understanding less so. Even in LSP, if you blindly apply the steps without understanding you are missing a big portion of your work as a facilitator. Now with AI, tools can write procedures in a snap but LSP can help us hold on to our “human magic”.

Three things AI does to a room and what LSP does instead:

• AI anchors the answer before you really thought about the question; with LSP everyone builds before we talk about anything;
• AI converges, every LSP model is different;
• AI invites deferring, taking shortcuts, while LSP carries a team to write their own principles for making decisions.

The LSP method asks from the facilitators to

1. understand the why: a change of mode from work to play makes an impact on cognitive, social, emotional, imaginative, emergent spheres.
2. Know the key questions to ask in the spur of the moment to identify what matters to the group, what’s the good (even the beautiful) thing to do? What resonates with the crowd? Feelings used as data transparently, and not as a source of bias to justify rushed decisions.
3. Stay an apprentice and take inspiration from your peers (cfr. Jacob Levi Moreno about spontaneity and professional role-playing). Repetition with feedback and debrief with a peer always helps; proximity is a tool, where ine facilitate and the other watches; frame first and fetch ideas later (especially when using AI).

Our work is to put the thinking first.