The subject of Artificial Intelligence (AI) and what it can contribute to media and broadcasting specifically is a topic so hot right now that pretty much by the time you finish writing an article about it, whatever you have written is behind the times. We can look forward to a feeding frenzy of ideas and applications about how radio can utilise AI to enhance, extend and expand our content horizons so I am not going to delve into that but rather focus on the implications of AI and longer-term considerations.
What got me thinking about this was a conversation with a friend who is about as far up the AI development food chain as you can get. Admittedly not necessarily in audio applications but certainly in the higher-level applications of this amazing and slightly scary technology.
My first question was “So what is possible?” to which the answer was “Whatever you can imagine.” This was quite a challenging response as I can imagine a lot, so I put to my friend the following.
“What about a radio presenter with the topical edge of Howard Stern, the comedic timing of Bob Hope, the humour repertoire of Joan Rivers, the debating skill of Margaret Thatcher and which sounds like George Clooney”.
My friend just smiled and replied “easy, so that’s all you want?”
Well, it is not in fact as “easy” as all that, but it is entirely possible if the boffins have access to enough of the right reference data and modelling parameters.
This prospect of engineering super AI personalities may seem fantastic, but it is comfortably within the remit of the technology. The question is not so much what is possible but where we draw the line. At present it seems the excitement about AI is moving well ahead of factoring the implications that may result.
A recent example of this was the German magazine “Die Aktuell” which some weeks ago promoted an interview with Michael Schumacher stating the interview would “answer the burning questions that the whole world has been asking for so long”. However, the interview was not of course, with the real Schumacher but rather written by AI on a chatbot. To the horror of the Schumacher family the magazine sold it to their readers as a “world first”. The magazine even featured a photograph of a smiling Michael Schumacher on the front cover promoting the fake story. The Schumacher family are of course taking legal action and many wish them well in that pursuit.
At this stage of the game no-one in their right mind can lay claim to knowing where AI is headed. It is potentially the most expansive technology ever created because at a point it transcends human influence and becomes self-learning and self-developing if designed and allowed to do so. It will not be a rabbit we put back into the hat easily because the commercial opportunities of AI will create such momentum and demand for the technology it will be a brave person who argues putting the brakes on and pausing to think for a while.
So it’s not so much a matter of debating the potential and limits of the technology but rather the implications of the technology.
In terms of broadcast some of the implications include:
Copyright: What protections if any, will personalities have to their voice, content, style and personal story?
Security: What protections will we need to prevent malicious interference, such as hackers accessing your AI traffic reports and turning them into a guide to Karma Sutra positions or a disgruntled employee deciding to leave you with a “special” message in the 7am news, read by your AI news anchor.
Competitiveness: What if at the flick of a switch your competitor can have an AI personality that is funnier, smarter, wiser and knows exactly what “less is more” is all about compared to the real presenter you have that costs a fortune, never shuts up and is difficult to manage.
Legal: What if a malicious interference leads to your station AI voice defaming someone or spreading false information. Where does a station sit legally with that? For some reason I think a defence of blaming the AI will not stand up in court.
Listeners: What will listeners think about it all? Will they even notice? Will there be a value placed on a real presenter or will they not care? Will it simply be about who entertains and informs them the best.
Presenters: What does it mean to real presenters when they are competing against more knowledgeable, funnier AI presenters who never have “bad days”, deliver a perfect sense of timing, don’t take holidays, never get sick, have no problem filling a 6-hour show on a Sunday, do everything they are told and are cheaper?
Radio’s Competitors: What does it mean to radio when our streaming competitors can replicate everything a radio station can offer with at least an equal degree of personality, content and localism?
The main message is that while we all rush towards what AI can do for radio, pause a moment to consider what it may do to radio. This is not something you can stop or ignore. Look for the opportunities, protect against the threats, avoid the hype and pause occasionally to think.
By the way, this was written by the real Wayne Clouten…not a Chatbot. Trust me, I’m a consultant.
By Wayne Clouten, BPR
This story first appeared on RadioInfo.asia