.author-name { display: none; }



Yesterday we spent some time exploring the ramifications of AI for marketing at Supercharged – Marketers and Machines, a conference bringing together the leading players in AI and technology for a day of bot chatter, AI-brewed beer and sarcastic supercomputers.

The day covered a lot of ground, but a few key themes surfaced:

Can consumers grow to love AI?

A subject raised by Nick Asbury, publisher of the brilliant Disappointment Diary, and surfacing again later on the panel – how can brands get consumers comfortable with an AI interface? And at a time when many brands are moving towards a spoken UI, has anyone checked to see if users actually want it? We text rather than call and hate hearing our own voices, especially when shouting our date of birth at a machine for the seventh time.

One of the key takeaways here was that small visual tweaks can give an AI interface a character, and playing with different personality types can add enjoyment to the user experience – but it’s tricky for many brands to apply.

A wisecracking financial management app might well make you pay more attention to your spending patterns, but do we really want to be sorting out broadband issues with a melancholy chatbot or ordering a curry from a sarcastic droid? At least these more extreme traits lead us away from another troubling personality issue - how many voice-interface AI default to the identity of a subservient woman – is this driven by research, sexism or laziness?

On the theme of bots with personality, Nick is the man behind @Botconference, a Twitter bot which live-tweets marketing jargon from fictional conferences. Here’s a taster:

“Do epic shit. Moment marketing laughs at digital nomads. The semantic web is the new disruption.”

So true.

Can AI make or save my company money?

A key point, succinctly put from Phrasee founder Perry Malm “The only reason you should be playing with AI is if it can make you money or save you money”.

Phrasee is a text-optimisation platform, which generate email headings and forecasts their success. They showcased 10 email headers generated for a pizza company, of which 9 were written by AI.

Sadly for all the non-robots amongst us, the AI open rate vastly outperformed the human-generated ones.

This came up again in the panel debate – will we reach a point of singularity, where there’s one email header to rule them all? And how does a brand maintain an identity in a world of personalised, optimised content?

Also on the horizon is the ability for AI to generate brand tag lines to compete with the current nonsensical crop, including Adidas’ “Impossible Is Nothing”, and Time Warner's “Enjoy Better”.

It turns out AI can also break grammatical rules humans don’t even know they’re following. 

The Twitter Bot @Taglin3r does just this, once an hour: here are some recent gems:

- Enjoy Reliability

- Tomorrow’s Oneness Matters

- More Efficient On Time

Where do ideas fit in?

Paul Lenaghan, Marketing Services Director at Nectar has seen success with AI – but only for optimisation, not content generation. He raised concerns about leaving AI to create and execute content unchecked due to its lack of empathy.

He also noted that AI can create all the content in the world – but people have a natural desire to own ideas.

Patrick Stobbs, co-founder of Jukedeck, a technology that uses AI to create music believes creativity arises from immersion, assimilation and recombination of ideas – all something that AI can do.

The panel broadly agreed that AI will be able to stir genuine emotion, pointing out that the Ferranti Mark 1, the first AI-enabled computer (built in 1948) was immediately tasked with creating love poems.

Will AI one day be able to generate authentically emotive art?

So, a lot of ground covered – but it’s probably best to leave the summary to @Botconference:More:

"Just work hard. The planet needs convergence. It's not about responsive web design. AI doesn't even exist any more."