UK artificial intelligence (AI) startup Phrasee has announced a $4 million funding round led by the Albion Capital Group. The investment will drive Phrasee’s ambitions for international expansion, which include scaling up operations in the company’s new US office.
Phrasee specialises in marketing copy – but that’s where the similarities with other agencies end.
Instead of traditional wordsmiths, Phrasee employs AI, along with developers, language technicians, and data scientists. The team combines a natural language generation algorithm with deep learning to provide marketing departments with brand-compliant copy, which has proven effective in direct, minimalist advertising campaigns.
Phrasee’s AI writes email subject lines, Facebook and Instagram ad copy, and push notifications – all of which have outperformed versions written by humans for the likes of Virgin Holidays, Superdry, and Domino’s, according to the company’s research statistics.
Adapting the role of AI in marketing
Traditionally, AI solutions involving natural language have been among the most challenging for researchers to develop. Even Google Duplex – though convincing enough in staged demos – can be caught out. Meanwhile, simplistic article-generation software is usually easy to spot when compared with a human writer.
But when the interaction doesn’t need to be an interaction at all, a significant complication is removed from the equation. The relatively brief copy that Phrasee specialises in – email subject headers, for example – is rarely more than 10 words in length. And the results appear to prove that AI has what it takes to push our buttons and persuade people to part with their cash – at least, according to Phrasee.
Parry Malm, Phrasee co-founder and CEO, has big ambitions for the company as it looks to gain a foothold in the US market. “From day one, our vision has been to transform how brands speak to their customers online. Having built awesome relationships with some of the most progressive brands in the UK, we’re now ready to take our AI technology and approach to the US market,” he said.
“The investment from Albion, and the experience it will bring to the board, will supercharge our global ambitions.”
AI-powered marketing and ethical concerns
We don’t need to look far to find examples of AI being used for data misuse and/or targeted advertising. For example, it was alleged today by Bloomberg that Google and Mastercard have a covert deal in place to track whether online ads lead to sales in physical stores, according to claims made by insiders. Mastercard and Google customers would have been unaware they were being tracked as the deal was never made public, says the report.
The escapades of Cambridge Analytica and the dawn of GDPR in the EU – if not yet in the US – mean that companies are under more scrutiny than ever to treat customer data ethically, particularly when it comes to analysis and marketing.
Meanwhile, a number of recent research studies have highlighted how susceptible humans are to coercion at the hands of intelligent machines. Both of these facts would, you’d imagine, put a marketing agency built on AI on shaky ethical foundations. But Phrasee has got out ahead of these contentious issues.
The startup was recognised by CB Insights as 2017’s Most Innovative AI Company and recently – like Google – launched an AI ethics policy. Among other things, the policy promises that Phrasee won’t use data to target vulnerable populations or promote the use of negative emotions to exploit people.
Indeed, those two standards alone would arguably put this AI-powered marketing agency on a higher ethical footing than some of its human-powered competitors.
Phrasee gets specific in its ethics policy: “It’s possible, for example, that a machine learning algorithm could identify bi-polar people about to enter a mania phase and then suggest you target them with extravagant product offers. Or, say someone has recently gone through a breakup, or worse, a bereavement, and people used AI to exploit their emotional state,” it reads.
“We do not believe using AI to these ends is ethical. We believe that even though a machine’s prediction may be right, that doesn’t mean we should use it to exploit people. We will NEVER use data like this.”
Of course, the implication is that someone else might.
Internet of Business says
In most discussions about AI, machine learning, and robotics, the assumption has often been that these technologies will mainly remove the drudgery from routine tasks, freeing up human beings to be more creative or to have their expertise enhanced by smart pattern recognition.
However, developments such as Phrasee’s technology suggest that, in the long run, creative careers will be far from immune from the spread of machine intelligence. Indeed, anything that is a non-core cost centre for enterprises is likely to be in developers’ crosshairs, given most organisations’ quest to remove costs, rather than actively make their businesses smarter.
Over the past few years, marketing, PR, and even journalism have felt the onslaught of AI and automation, with the introduction of systems to automate the production of press releases, automate the sub-editing process at News UK via AI-powered headline-writing and copyediting (a senior figure at the company told Internet of Business editor Chris Middleton), and systems at other providers that generate news stories from press releases (all known to Internet of Business).
Meanwhile, all of these industries are being enhanced by AI tools as well, via cloud platforms such as Box, Salesforce, Microsoft, and others.
So the key for all sectors is clear: employees, even longstanding expert professionals, need to develop new, deep, enhanced, and/or transferrable skills in order to keep one step ahead of the machines.