Questions on Questions
Anthony Howcroft interviewed by Holly Self, VP Marketing at SWARM
Anthony, you didn’t mention how difficult it would be to ask you questions after reading your user’s guide - I’m sure this was all part of your plan! I went back through your book and re-wrote my questions multiple times and I’m still not totally happy with them. Anyway, here goes…
Questions, really? A book about questions? How much time can one person spend on questions?
Well, I didn’t realize quite how big a subject it was until I started the project. I had a notebook where I could jot down ideas, and since the book was very much a background project, over a year or two I just kept adding ideas as I encountered different situations or thought of some other aspect. When I went and looked at the whole list, I got my first inclination about the scale of this topic, because questions are integral to everything we do – how we go about our jobs, build relationships, learn new skills, and even in areas where you wouldn’t expect to find them – like music, or jokes. I probably had enough content for a book twice the size and had to cut down the material so it wasn’t overwhelming. The more I researched, the more amazing facts I uncovered - I hope readers will go on the same journey of discovery that enthralled me.
What compelled you to write the book, especially a user’s guide?
It came about by accident. I was asked to present at a Microsoft conference in South Africa, and rather than talk about our product and all the great things it could do, I thought I’d discuss how our customers could use the product to make a difference, if they approached it in a certain way. By chance, in the previous week, I’d managed a huge personal best in my running, thanks to a great question that a friend had asked me. I began writing a presentation on the flight – and it’s a 14 hour flight from London to South Africa, so I had plenty of time – and those two themes combined. As I thought more deeply about the subject, I noticed that my most successful customers were the ones that had asked me good questions during the sales process, and they’d subsequently used our product to ask great questions of their business. I based the presentation on that theme of questions, and got a really positive reaction from the conference audience. Afterwards, I couldn’t stop thinking about this, and decided I would learn more about questions, and document the process in a book. It wasn’t really intended as a user guide, it was really a self-help journal for myself.
What did you find most interesting in researching your book?
The section on non-human questions was probably my favorite – learning what happens when we teach other species to communicate with us, and thinking about how AI is evolving. To see that computers can already pass the Turing test (and fool us into thinking they are human), or to explore the questions we might ask dolphins, or how aliens might consider our planet - that was a lot of fun, and for me it shone a light on the critical element that make us human – not opposable thumbs, or vocalization, which are traits shared by many other animals, but remarkably, questions.
Oh, I was also blown away by the Elizabeth Loftus research that demonstrated how the questions we get asked change our memories. That’s profound. I’m still thinking about that, and the implications.
After the AI revolution, what will be the role of humans? Will we become overpaid order-takers or will new, advanced specialization happen?
We don’t yet know when a super-intelligent AI system will arrive, but when it does it may happen very fast. The intelligence gap between a future AI system and a human could be as wide as that between an earthworm and a human, which means that we really can’t conceive of how the AI will regard us, as it will be operating at such a different level to ourselves. In the best case, the AI system is benevolent, and helps us solve problems that have plagued humans for centuries: disease, war, famine, to usher in a new era. There are obviously darker scenarios, which might resemble the Terminator movies, but a more likely outcome is that the AI will largely ignore us – how much thought do we give to earthworms? There’s a lot of work on AI & ethics right now, so that we can try to embed a positive set of values in our future AI.
Given how terrible humans have proven on predicting the future, I’m not going to stick my neck out here! What I will say, is that before we reach superintelligence in AI, we will be increasingly sharing the workplace with AI-enabled colleagues and partners, which I believe is a great thing. Technology can help us achieve fabulous results in many areas. At SWARM we’re helping to reduce food waste, lowering companies carbon footprints by minimizing un-necessary truck journeys, while saving time and money for organizations – and all of that is with the help of AI, but with the goals being set and controlled by people. One of the key discoveries I made in writing Questions, was that humans seem to have a unique capacity to formulate questions, and AI is especially good at helping us answer them. That combination is incredibly powerful and will reap huge benefits for us as individuals and society.
When you (personally) have a question, where do you turn for the answer? An individual, a crowd, a swarm?
Well, I guess it depends very heavily on the type of question. If I want to find a place to stay in the South of France, I’m probably going to be hitting the internet, but I might not do that to answer a question that requires a certain level of expertise. I would be more likely to ask a trusted advisor, somebody I know with experience on the topic. I have to say, one of my most common places to search for an answer is Wikipedia, which I think is hugely under-rated. It survives on donations, and I’d encourage anyone who uses it to support the site.
How do we decide who to ask? That sounds like another section I should add to the book! It really does depend on the context. Crowds have been shown to be great for estimation – solving the classic fairground problem how many jellybeans are in this jar? Swarms are a new way of answering questions, and you can think of it as a crowd of AI computers. They are great at solving problems with large numbers of variables, problems which have got too complex for a human to comprehend and solve in a realistic timeframe. Things like logistical problems, moving potentially perishable products through a transportation network at minimum cost, even as the system is disrupted with floods, fires, or pandemics. As we add more data and variables to support our decisions, and need answers at an ever faster pace, I believe we will increasingly turn to swarms of AI agents to help.
In a group setting, which type of question asked lets you know the person is friend or foe?
The questions people ask are revealing, and often show their underlying mental model of the world. I don’t think there is a specific type of question that reveals intent – in terms of open or closed, rhetorical and so on, but if you consider why someone asked that question from the many they could have chosen, you will have a good idea on whether they are friend or foe. Imagine if you tripped over a cable in an office and fell over. You’d hope that people would be concerned about your wellbeing, and would ask Are you OK? but if they said Didn’t you see the cable? then the subtext could be that they think you’re an idiot.
In group situations, the questions we ask are often based on the role we play. In a sales meeting, for example, a technical expert would be expected to ask questions about the technology under review and people might be surprised if he began to delve into the pricing model and the purchase terms. If you didn’t know that person’s job title, you could still figure out their role, and even how positive they felt towards the technology, based purely on the type of questions they asked. We all know that intuitively. They say eyes are the windows of the soul, but it seems like questions are the door to the mind.
What is the most important question a human can ask?
Were you reading the book section on killer questions? Because that’s a tough one to answer! There are various thoughts that popped into my mind, such as How do I live a happier life? or How do I make a difference? But of all the questions, the one that seems to me to be most critical is What should I do next? We all live in the present, and our next action determines everything about our life. It’s never too late to alter direction, make a change, shift an attitude, to say yes instead of no, and vice versa. It seems to me that deciding how we spend our time is the most important question.
What’s next for you? How are you applying your research to work and life?
Caught by my own question! I’ve been working on this book for many years, so the lessons I’ve learnt are embedded in the way I think and interact with the world. I’m no master, though – questions are something we can all get better at using.
At SWARM, we are using the research to drive the next generation of easy problem-solving. Let me explain. We frequently talk with executives at food companies, trying to understand challenges in their supply chain. These people are managing areas like logistics and are incredibly busy handling operational issues. Asking the right questions to find out how we can help each other is critical, since we don’t get much time together. Today, SWARM has a modeling tool that captures customer challenges and problems, in such a way that we can use this to find and deploy a solution. Often, we do the problem-modeling for customers. We have begun work on the next generation of this modeling tool, and are intentionally looking at the research from Questions, to make a step-change in the process of problem-identification & solution-mapping.
Imagine if instead of building a detailed model of events, goals, constraints, and process steps, what if you could share some data with SWARM, and discuss your challenge for 30 minutes with an AI agent, who would define the problem shape for you and immediately provide a fully working solution? That’s where we’re driving the research; to help people solve problems far more easily. There’s a multitude of fantastic optimization algorithms and machine learning solutions being produced by academics, that can solve most typical business problems – but it’s really hard to find, understand, and deploy them. Our goal is to make this happen automatically, and rapidly – and for the solutions to keep improving, too. We believe good questions are fundamental to this approach.
That sounds exciting! Where can I buy the book, and how do I find out more about SWARM?
The book is available now on Amazon as a paperback or a kindle ebook, here. If you want to learn more about SWARM, and how it could help your business, feel free to reach out to me directly. My email is anthony@swarm.engineering