Only Dead Fish

When Robots Bluff

Accepting the (current) limitations of machine learning (and some recent hiccups), it's still fascinating how AI capability is developing so fast despite the field being far from new. The 'AI beating humans' narrative (as Tom Chatfield talked about at last year's Google Firestarters on the theme of AI - describing it as the ‘usurpation narrative of human-machine interactions…a creation is pitted against its creators, aspiring ultimately to supplant them’) is not always that helpful. But still, the landmarks along this storyline (AI beating humans at Jeopardy and more recently the famous victory of Google Deepmind’s AlphaGo against GO world champion Lee Sedol) are quite compelling.

So now we have the AI system (developed by Carnegie Mellon University) that has just accumulated $1.7m worth of chips over twenty days playing against four of the world’s top professional poker players. What was particularly remarkable about AlphaGo's victory over Lee Sedol (which I remarked on at the time) was the way in which AlphaGo won, playing in highly unexpected and creative ways, with moves that left even its creators 'pretty shocked'. Wired described how:

‘…with its 19th move, AlphaGo made an even more surprising and forceful play, dropping a black piece into some empty space on the right-hand side of the board. Lee Sedol seemed just as surprised as anyone else. He promptly left the match table, taking an (allowed) break as his game clock continued to run.’

The commentator at the time noted how Sedol appeared to have trouble dealing with a highly unusual move that he had never seen before. It was a real moment in the development of AI/human relations.

And perhaps we've just passed another one since what’s again interesting about the poker example is that on its way to victory, the AI succeeded in out-bluffing its human opponents. Poker, the FT piece notes, tests different mental muscles since it involves 'strategising using imperfect information in a way that is more akin to the real world'. The AI started not so well, but then focused on fixing its own weaknesses and filling the holes in its strategy as it went. It was able to use an additional feedback loop to respond in real time to the human activity. The implications are that AI can take other imperfect information situations and derive a good strategy for that situation.

Another new milestone.

Image courtesy

The Long Nose of Innovation

Thanks to Faris for pointing at (in his excellent post on the paradox of influencer marketing) Bill Buxton's idea of the 'long nose of innovation':

'My belief is there is a mirror-image of the long tail that is equally important to those wanting to understand the process of innovation. It states that the bulk of innovation behind the latest "wow" moment (multi-touch on the iPhone, for example) is also low-amplitude and takes place over a long period—but well before the "new" idea has become generally known, much less reached the tipping point.'

Much of what Buxton says chimes with points that I make about the innovation process in my book. There's a general fetishisation of the early stages of innovation, most notably ideation. Much of the hidden assumptions or mistaken beliefs around how great innovations happen focuses on the narrative of a lone inventor who, with a sudden spark of inspiration, originates an idea that takes us forwards in one great, sudden leap. Yet, as Stephen Johnson demonstrates well in his great history of innovation Where Good Ideas Come From, the best breakthroughs often come from groups of people and the collision of ideas over time, meaning that connection is one of the best ways to empower ideation.

Buxton makes the point that many breakthrough ideas (even in fast-moving sectors like technology) take a long time to gestate, bouncing around between university research teams and corporate innovation labs and so on. The mouse, for example, which we think of as a breakthrough innovation that Steve Jobs copied from Xerox PARC for the Macintosh, had actually already been around for almost twenty years by that point, and in the end took thirty years to go from original demonstration to mainstream.

I talk in the book about the need in innovation to see the trajectory, to pay attention to fundamental behaviours or needs in the context of developing technology trends, but in particular to know when it's right to jump to the next S-curve before the existing one plateaus. This might appear at once difficult and dangerous but often the signs are there if you look hard enough. As Buxton says, any technology that is going to have significant impact over the next 10 years is already at least 10 years old. Whilst they might not be mature or their application fully demonstrated, the basic concept has likely been there for a good while.

The importance of understanding how to focus (and why it's important to focus) on each key stage of the innovation process (I use Schumpeter's definition as a way of expressing this) is also something I emphasise in the book. Arguably the biggest challenge with innovation comes not with ideation (most companies in my experience are not short of great ideas), but with the harvesting, application, commercialisation and scaling of those ideas. Buxton notes too, that innovation is not about alchemy or invention, but the bulk of the work and the creativity is actually in the augmentation and refinement of that idea ('The heart of the innovation process has to do with prospecting, mining, refining, and goldsmithing. Knowing how and where to look and recognizing gold when you find it is just the start. The path from staking a claim to piling up gold bars is a long and arduous one.'). We need, he says, to reorient investment and prestige to reflect this. 

Amen to that.

For more like this, and for exclusive content related to my upcoming book on Building the Agile Business, you can sign up here.

Image source

What is Agile Marketing?

Another one of those popular yet ill-defined phrases this isn’t it? When I hear people talking about 'agile marketing', they’re mostly speaking about fast-turn-around, tactical, opportunistic messaging which (when done well) can amplify in the short-term but (when done badly) can be truly awful. But I think there’s so much more to it than that. So here (for what it’s worth) are the three categorisations that for me capture what agile marketing is really all about:

1. Responsive: the kind of reactive or topical activity that I mentioned above. The impact of this type or marketing is IMHO often over-rated, though it can be a useful tactic to have as part of a much wider approach to content. I’m not a big fan of the term 'brand newsroom’ but to make this work properly you do need to have the right people coming together at the right time, empowered with the right sign offs and tools.
2. Test and learn: something altogether different but fundamental now to marketing operations. Test and learn is not new (consider classic direct marketing approaches) but what is different is the scale, breadth and type of testing that is now open to us. There is huge opportunity in this. Not least from the growing automation of marketing operations. The thing I see in most marketing teams is an episodic or one dimensional approach to testing. But it's about testing all the time in multiple ways. It's about developing a true testing culture where learning is embedded into practice and process so that it happens continuously and at scale. But automation has to be deployed carefully. And we need absolute clarity around the fundamentals (customer insight, strategic direction, brand vision, positioning and voice), and about what can change, and what can't. Then it's about establishing fast learning loops to enable us to stay close to customer need, be adaptive and optimise well. I like to think of this as designing with vision, but optimising with feedback. Test selection should focus on giving us maximum learning in the simplest ways. But there is nowhere near enough testing being deployed in most marketing teams
3. Experimentation: related to this is the idea of agility through constant experimentation. The best way to think about this is to recognise the difference between testing around marginal improvement (in others words optimisation), and testing in order to deliberately learn new things that can lead to a breakthrough or a new direction. Both are necessary but test selection (controlled experiments designed around learning) is again key. Test and learn is not an opportunity to do dumb stuff.

That's my definition.

Post of the Month - January 2016 - The Winner

Well, a real run of votes has meant that Lazar Dzamic's post on behavioural economics has won the vote. In fact it completely ran with it which, in a way, is a bit of a shame as it's somehow better when there's a more even spread of votes. Nonetheless, well done to Lazar, and the post is in the Hall of Fame. Thanks for taking part.

Post of the Month - January 2016 - The Vote

Thanks for the nominations. So our vote this month is between:

The Schedule and the Stream by Matt Locke

How I Got My Attention Back by Craig Mod

Planners Need to Get Out of the Echo Chamber by Richard Huntington

Are you a cowboy or a farmer? Managing the tension between the pioneering spirit and the need to cultivate the land by Jim Carroll

Art vs Science and marketing in a post-truth world by Jon Bradshaw

Should Behavioural Economics in marketing be regulated – or hyped-down? from Lazar Dzamic

And you can vote below: