Who asked for AI? Why its future is more boring than you think
Photo: ThisisEngineering
Henry Ford famously once said, “If I had asked my customers what they wanted, they would have told me a faster horse”. This quote is often trotted out by ‘innovators’ when defending a new piece of technology; consumers do not know that they want it, because they don’t have it yet. Just you wait and see.
Of course, a car essentially is a faster horse. It accomplishes what people need – getting from one place to another – faster than a horse, more comfortably, and (in the best cases) with style, but it still is there to do exactly the same thing.
The smartphone was similar. While unique from its predecessors and ultimately reshaping society as we know it (much like the car), at its core, it addressed things people were already doing and did them ‘better’. Instead of carrying around a handheld phone, a digital camera, an MP3 player, a map, and a book, you could carry one (small) thing that took pictures, called people, played music, gave you directions, and could look up reading material (or anything else entertaining) online. While new use cases have since emerged (from Fruit Ninja to TikTok dances), the smartphone’s function can be distilled down to those core needs of communication, entertainment, and utility, with a little creative flair, offered at the utmost convenience from your pocket.
At the core of every success or failure for new technology is the crucial question: does it address an existing need of its intended audience? And do so better (faster, cheaper, more efficiently, more pleasantly) than existing alternatives, if there are any?
The widespread enthusiasm for AI, which has infiltrated every email newsletter, company announcement, and conversation over the last 18 months, does not necessarily bear out this line of inquiry. While AI itself can have myriad implications and uses, the cultural fixation is on purely generative implications – the Soras, Udios, and ChatGPTs of the world. But market-level enthusiasm has not (yet) manifested entirely into reality. From faked demos to biassed outcomes, and of course the ongoing battles over the future of copyright itself, premature launches and a desire to jump on the hype train to please stakeholders have resulted in retroactive adjustments and a pervasive unreliability.
Nevertheless, there are needs that generative AI is managing to address, for better or worse. ChatGPT is used by more than half of 16-19-year-olds, who seem keen to have it do their homework (user need case, check; educational institutions, floundering). Suno and Udio produce music that is perhaps not good enough to make the charts, but is definitely good enough to replace the music in 2 a.m. pizza ads or soundtrack the dramatic buildups in an episode of Married At First Sight. Paying for music is a problem for media companies, and generative AI could cut those costs, thus ‘solving’ it. Toys R Us recently released an ad made with OpenAI’s Sora, ultimately addressing many of the same needs on the business side: lowering costs to produce ads.
Featured Report
Analogue revival A cultural pendulum swing
From the dumb phone trend to a vinyl revival, analogue has been back on the rise in the cultural zeitgeist. This could shake up the digital ecosystem and unlock potential opportunities for entertainment...
Find out more…Relying on generative AI – and the costs it can cut, from production companies and music artists to employee headcount – is a tempting short-term benefit for many companies. However, with its ongoing unreliability, this might not bear out successfully in the long term, and is ultimately reliant on consumers finding added value in the output. A customer support chatbot that just sends them back to the FAQ page isn’t helpful. More content being produced for cheaper does not benefit them, when they are already drowning in options on streaming, and largely want to avoid ads.
Consumers themselves do not necessarily find huge value in purely generative AI, aside from having it do their homework or write their emails. Our latest report finds that they are more interested in AI tools that can help with modification and streamlining processes, but not to do the entire creative work for them; the IKEA effect comes to mind. In the words of the now-famous tweet by author Joanna Maciejewska: “I want AI to do my laundry and dishes so that I can do art and writing, not for AI to do my art and writing so that I can do my laundry and dishes”. If the goal of technology is to make people’s lives easier and give them more leisure time to spend on their passions, then purely generative AI is not doing a very good job, given that it seems to be cutting out the time ‘needed’ for leisure and creativity, and upping profit margins for businesses at the expense of consumer income.
The longer-term need for AI to solve, then, is ‘doing the (proverbial) dishes’. Unfortunately for headlines, this tends to be rather boring. Fortunately for the companies building those products, however, they never go out of fashion. A fixation on the flashiness of generative AI to replace jobs (and ultimately undercut quality) misses the massive opportunity of pragmatic assistance, which can be extended to everyone.
Modification tools that help make videos better is more appealling to creative consumers than a program that just generates the video for them. Independent artists who struggle with doing everything themselves do not need Udio to make music for them so they can focus on marketing and interpreting legal contracts. Rather, they need tools to help them understand contract terms, and help schedule and optimise their marketing collateral, so they have more time to make music. AI’s ability to ‘smart search’ massive documents and databases without knowing the precise phrasing of what they’re looking for is incredibly useful for researchers, lawyers, and medical professionals. In-real-time language translation can make a huge difference for reporters seeking balanced perspectives or police departments dealing with foreign visitors in distress. Information syncing (e.g., calendars) across platforms that are not otherwise interoperable is helpful for long-time web users with scattered digital footprints.
These things are not terribly exciting, and certainly won’t make the latest headlines. Infrastructure rarely is – but ends up being the most important thing of all, unremarkable when working and most notable for its absence, rather than simply being a flashy value-add easily discarded. We shouldn’t pay all of our attention to the Soras and Udios, and instead focus more on those plugging away in the background, doing what technology is supposed to do: making our lives qualitatively better and a bit less stressful, not the other way around.
To dive deeper into the opportunities for entertainment and AI, check out our latest report, The state of music AI: The consumer opportunity lies in modification, not generation.
The discussion around this post has not yet got started, be the first to add an opinion.