Getting AI in perspective

I can’t help but note the irony in quoting this by Barry Adams (from LinkedIn):

The news industry is too focused on AI.
Every industry event is dominated by AI talks and tracks. Every media pundit column is about AI (and yes I've contributed my fair share). Every talking point about the future of news starts with AI.

Yup. Here I am, a media pundit, delivering another column about AI. So, yes, I know, it seems like I’m adding to the problem right now. But I swear I’ll reverse that before we’re done.

Anyway, enough self-flagellation. Let’s get back to Barry:

With all this talk about AI, there isn't space to talk about everything outside AI. You know, the actual stuff that matters. The unique value a publisher can bring that binds audiences to you, that builds loyal readerships, that immunises you against whatever the tech bros dream up next.

Yup. Lots I agree with here. Admittedly, a cynic might note that somebody whose best known area of expertise is SEO might have a vested interest in downplaying AI talk, given the risk it poses to search traffic. But this particular cynic rather agrees with him, plus is also inclined to think that the hype about “Google Zero” is just that – hype. It would take a profound change in behaviour for traffic from search to be eliminated entirely, even if AI overview got good enough to be entirely trustworthy – which they’re a long way from being. Significantly diminished? Sure. Already happening. Completely eliminated? I doubt it.

The tech over-correction in journalism

The problem leading to this AI over-saturation isn’t difficult to see. As I’ve noted before, today’s managers are people who watched the last generation of managers completely botch the transition to the digital age, by being too sceptical and slow to embrace it. They are now, as is so often the case, over-correcting. They’re being too credulous about the potential of a new technology, too willing to throw all their attention at it.

And they’ll end up regretting it just as badly. I’d bet a serious amount of money on some of today’s AI experts finding themselves trying to quietly erase some of their more ebullient prognostications from the record in a few years’ time…

This, to be fair, is hardly an issue unique to journalism. We are seeing classic bubble-like behaviour, with companies across all sectors trying to re-orient themselves around this emergent tech. Watching the tech companies’ commitment to environmental goals crumble in the face of AI’s insatiable demand for energy and water is teaching us a lot about how seriously you can take their word on anything. And is a mix of infuriating and depressing.

The audiences do not love AI

But what is unique to journalism is that we have clear signals that audiences are really not happy about AI-generated journalism – and that they’re getting more unhappy over time. Reuters Institute Research, I choose you!

See?

The great irony here is that everyone wants to use AI, but pretty much nobody wants to consume its output. That dilemma will haunt all content creation businesses for the foreseeable future, as will the overblown hype about the technology, which is an open secret even in Silicon Valley.

And that loops us nicely back around to Barry’s post:

AI is here and yes it is having an impact, but we should talk less about it and more about what we can do outside AI to grow our brands, grow our audiences, grow our presence on the web and beyond.

Ah, yes. Audiences. A much more important word starting with “A”.

Putting the tool into perspective

I was running my Audience Engagement Strategies course this afternoon, and it was refreshing to be able to focus on just that: starting with the audiences, and then working out which tools to deploy to give them the journalism they want, when they have time for it, in the format they need it in. And when you view things in that context, AI is just another tool in the toolbox, rather than the centre of the universe.

Some of the rest of our competition is already ahead of us on this. I attended an excellent talk at NEXT in Hamburg earlier in the year, about the future of the website. The marketing folks in the orbit of Accenture Song were talking about how critical your own website remains in building relationship with customers, even in the AI age.

They even discussed the idea of websites operating at two levels:

• one level for human consumption
• another for the AI crawlers to ingest

In the months since, that has been playing on my mind. There must be some variation of that approach that would work for journalism businesses – but I haven’t cracked what it would look like just yet. But the idea won’t leave my mind.

But, my friends, when marketing people are doing better than we are on keeping the focus on the audience, not the technology, then we are going very seriously astray.

Asking the right questions about the capabilities of AI

So, it’s time to rewind. It’s time to put on our journalist’s big person pants, and start applying the same level of rigour to analysing the claims of the tech bros we would to any other subject. Yes, AI will be a tool in our toolbox, and a powerful one. I am not an AI sceptic. I'm teaching my first course about it right now – and will be doing so in future. Nevertheless, it’s not going to be the centre of the universe, and we shouldn’t let it monopolise our attention.

We made this mistake 15 years ago. We handed our audience relationships over to Google and Facebook, and have been trying to undo the damage that did for the last seven years. Could we, just for once, actually learn the lesson of the past, and get some sort of perspective on this technology, before we shit all over some hard-won audience relationships?

Here’s how to get started (and it’s a fun game to play). Next time you see an “AI expert“ listed on a conference schedule, check their posting history. Were they a Metaverse expert a few years ago? And a blockchain/crypto one a few years before that? If so, you can safely ignore them. They have no more insight into the way the world is going than you do. They’re just really good at spotting today’s trendy topic and persuading conference organisers to cough up for a speech on it. Not only that, but they rarely have any useful insight, but perform the questionable role of making senior executives feel confident that they can go back to the office and tell people what to do.

The snake oil was never so slick, so well-packaged, so well-supported by a beautifully designed slide deck.

I know some of these people of old. And I remember when they were e-commerce gurus, or social media experts. They’re not and never were, They’re just expert surfers of the trendy conference waves.

Seek the voices of AI experience

On the other hand, if you look at a schedule and see that they've been working for years with AI in a particular context relevant to your work, then by all means pay attention. One good example were the people from O’Reilly Media at the Future of Media Technology conference a couple of months ago.

Look, there’s plenty of great work happening in AI and journalism. Go and look at this post from Metro’s Sofia Delgado (an audience expert, of course) on LinkedIn:

My favourite part of #JournalismAI festival earlier this week was experiencing a space with global diversity running through the core of it. I may be a little biased, but the case studies speak for themselves… the global south, with its melting pot of needs, creativity and lean teams, is coming up with highly valuable solutions.

Yeah, that’s more like it. A whole list of practical experiences in a real context, not the normal stage bullshit. Let’s keep the conversations about AI real, focused – and in their place.

There are plenty of other things we should be talking about, too.

And all of those start in one place: audience needs. Now, shall we see if I can get to the end of the week without posting about AI again?