{"id":3734,"date":"2024-08-16T09:14:40","date_gmt":"2024-08-16T08:14:40","guid":{"rendered":"https:\/\/incite.strat7.com\/?p=3734"},"modified":"2025-06-02T11:42:32","modified_gmt":"2025-06-02T10:42:32","slug":"getting-out-of-book-the-worthy-rival-challenging-researchers-to-have-more-original-conversations","status":"publish","type":"post","link":"https:\/\/incite.strat7.com\/en_us\/blog\/getting-out-of-book-the-worthy-rival-challenging-researchers-to-have-more-original-conversations\/","title":{"rendered":"Getting Out Of Book: The \u201cworthy rival\u201d challenging researchers to have more original conversations"},"content":{"rendered":"

This article is about supercomputer chess, small talk, and the rise of synthetic respondents in market research. To understand the connection between these unlikely bedfellows, I need to briefly back up.  <\/strong><\/p>\n\n\n\n

The Most Human Human<\/a><\/em> is one of the top three most thought-provoking books that I have read in the last five years. My copy has hundreds of sentences underlined and handwritten notes scrawled on almost every page. Published in 2011, it still feels ten years ahead of its time.  <\/p>\n\n\n\n

The author Brian Christian, an American poet and computer scientist, entered the Loebner Prize<\/a> \u2013 the world\u2019s most famous Turing Test. In the competition, teams of AI programmers try to persuade a panel of judges that they are talking to a human through a digital interface. The team that successfully convinces the most judges wins the Most Human Like Computer prize.  <\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n

Christian entered the competition as a human blind, speaking with the judges as \u2026 himself. He was eyeing a different prize: the Most Human Like Human. That is, the confederate who most convincingly persuades the judges that they are speaking with a flesh-and-blood person. But how might you convince a skeptical judge of your humanity through an online chat? What is it that makes real conversation truly spark? There are endless articles on optimising large language models, but very little guidance on how to be human. The book explores what it means to be human in a world of artificial intelligence.   <\/p>\n\n\n\n

I want to share one insight from my favourite chapter called \u201cgetting out of book.\u201d The concept of being \u201cin book\u201d comes from chess: the set of opening moves that all serious players memorise and study. The list of possibilities is finite and, compared to later moves, relatively predictable. It is sometimes said that a game doesn\u2019t meaningfully begin until the players are \u201cout of book\u201d, into the realm of creativity and complexity.    <\/p>\n\n\n\n

Incidentally, the chess match between Gary Kasparov, the greatest ever player, and IBM\u2019s supercomputer, Deep Blue, was lost in just 19 moves\u2014that is, while still in book. In the cultural commentary<\/a> that followed, Kasparov\u2019s loss was discussed as a blow for the whole human race, but he didn\u2019t buy into it. Kasparov said that while he lost, IBM didn\u2019t win. Because he was still in book, because he made a procedural error, the game never really got going. Initially this may seem like an elaborate defense mechanism, but he has a point. To understand why, we need to think laterally.  <\/p>\n\n\n\n

Christian makes a brilliant connection between playing \u201cin book\u201d and small talk. It might be said that real conversation doesn\u2019t take flight until we move beyond water cooler weather commentary. When you think about it, isn\u2019t there something a little robotic and pre-programmed about these conversations, with each person knowing and executing the shared social code? Isn\u2019t it predictable, with a limited set of opening moves? Conversation, like chess, the real heavyweight variety, only gets going once we are out of book.   <\/p>\n\n\n\n

What does this mean for market researchers? <\/h3>\n\n\n\n

Recently I\u2019ve been thinking about \u201csynthetic respondents\u201d, which are virtual representations of a key group. It is eerie and disturbing that a bank of historical respondent data can predict and match real-life survey data with 95% accuracy<\/a>. Eerie and disturbing \u2026 and impressive. It is our industries\u2019 equivalent of The Most Human Like Computer, a worthy achievement.<\/p>\n\n\n\n

Indeed at STRAT7 we are actively building solutions that help bring users closer to representations of their segments with AI<\/a>. So I don\u2019t doubt that this provides value, and yet I can\u2019t help but wonder what the high level of predictability says about the nature<\/em> of our conversations. <\/p>\n\n\n\n

In a world of advancing AI, how do we have more original conversations? How do we get out of book?  <\/strong><\/h2>\n\n\n\n

1 \/ Be Human <\/strong> <\/p>\n\n\n\n

On the one hand, thinking about people as consumers is necessary. We are never briefed to know the whole human person, however interesting that brief may be. And yet there is something dehumanising in the way we reduce the human animal to a \u201cconsumer\u201d. I picture an open-mouth Pac-Man mindlessly swallowing little yellow dots. Is it any wonder that \u201cconsumers,\u201d narrowly defined, are predictable? Thinking of consumers as people, situated in a broader context, with hopes and aspirations above and beyond the project objectives that nonetheless influence their behaviour, is a sensibility that inspires more insightful questions and elicits more unexpected answers.  <\/p>\n\n\n\n

2 \/ Be Curious<\/strong><\/p>\n\n\n\n

One of my favourite warm up exercises in online focus groups is \u201cshow and tell.\u201d I ask people to bring one object that lights them up. The transformation\u2014in energy, demeaner, aliveness<\/em>\u2014as people talk about their peculiar \u201cwormhole of fascination\u201d<\/a> is instant. And, most importantly, insightful. Understanding what matters to people and why will always teach you something. If you\u2019re ever at a BBQ or a party and the conversation begins to flag, ask someone what they are excited about right now. I have never once heard a dull answer.   <\/p>\n\n\n\n

3 \/ Be Bold    <\/strong> <\/p>\n\n\n\n

Part of the reason we have small talk is because it provides safe, neutral territory. Real talk requires risk. It becomes more personal, more vulnerable<\/a>. Clearly there are safeguarding considerations in research design, but getting out of book requires some courage both in asking and answering questions. One of my favourite qualitative exercises is a \u201cletter to your younger self.\u201d Sometimes these answers border on cathartic and therapeutic, but knowing someone\u2019s retrospective vision of life is often surprising and frequently moving.  <\/p>\n\n\n\n

A worthy rival <\/strong> <\/p>\n\n\n\n

It is remarkable that today\u2019s synthetic respondents predict actual respondent answers with 95% accuracy, but it is worth asking what percentage of those conversations are \u201cin book\u201d, and what we can do to get out.    <\/p>\n\n\n\n

I believe we should view synthetic respondents and AI more generally in the spirit of a \u201cworthy rival\u201d\u2014that is, an adversary that pushes us to be better. If we are asking questions with predictable answers, it behoves us to be bolder, more curious, and more human.  <\/p>","protected":false},"excerpt":{"rendered":"

This article is about supercomputer chess, small talk, and the rise of synthetic respondents in market research. To understand the connection between these unlikely bedfellows, I need to briefly back up.   The Most Human Human is one of the top three most thought-provoking books that I have read in the last five years. My copy has […]<\/p>\n","protected":false},"author":3,"featured_media":3769,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[7],"tags":[],"class_list":["post-3734","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-blog"],"_links":{"self":[{"href":"https:\/\/incite.strat7.com\/en_us\/wp-json\/wp\/v2\/posts\/3734","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/incite.strat7.com\/en_us\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/incite.strat7.com\/en_us\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/incite.strat7.com\/en_us\/wp-json\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/incite.strat7.com\/en_us\/wp-json\/wp\/v2\/comments?post=3734"}],"version-history":[{"count":2,"href":"https:\/\/incite.strat7.com\/en_us\/wp-json\/wp\/v2\/posts\/3734\/revisions"}],"predecessor-version":[{"id":3768,"href":"https:\/\/incite.strat7.com\/en_us\/wp-json\/wp\/v2\/posts\/3734\/revisions\/3768"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/incite.strat7.com\/en_us\/wp-json\/wp\/v2\/media\/3769"}],"wp:attachment":[{"href":"https:\/\/incite.strat7.com\/en_us\/wp-json\/wp\/v2\/media?parent=3734"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/incite.strat7.com\/en_us\/wp-json\/wp\/v2\/categories?post=3734"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/incite.strat7.com\/en_us\/wp-json\/wp\/v2\/tags?post=3734"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}