Where AI belongs in entertainment

Three years of AI as an everyday companion#

About three years after ChatGPT arrived, it has become clear how deeply AI has worked its way into everyday life. For better and for worse, it can no longer be ignored. For some people, AI itself is becoming something close to a human connection: always there to talk to, with no language barrier, filling the lonely moments. In that sense, AI really has become a new kind of conversation partner.

In entertainment — behind the scenes and in direct contact with fans — AI could change a great deal. It can give creators and staff room to breathe, and become an important way to make experiences more personal, more engaging and more meaningful. That is, if it is used for the creators and for the people enjoying what they make.

Chatbots, smarter ticketing and all kinds of operational tools can simplify how things run dramatically. They make customer support smoother and let staff focus on creative, person-to-person work. For me, the value of automation goes beyond saving time: it makes possible projects and ideas that time and resources used to rule out. That is the most interesting part.

What I find disappointing — and not only in entertainment — is how many services simply wrap a ChatGPT prompt and sell it as a product, and how many companies pay for them. That was never going to produce much. Yet the conclusion becomes “AI just doesn’t work for our business,” and most of these efforts seem to end within a few months.

Buzzword AI, and prompts sold as products#

Plenty of management teams have reacted strongly to whatever word was trending — NFT, web3, crypto, stablecoin, metaverse. Organisations that chase buzzwords seem to be repeating the pattern with AI. What I fear most in entertainment (and I am genuinely glad the company I work for understands this) is forgetting that the audience is people. Whatever the tool, whatever the workflow, it means nothing unless people end up finding it fun, or loving it.

“The metaverse is the future,” we were told, and many executives poured years of money, assets and time into it — while users kept asking what it did that VRChat didn’t, or what the metaverse even was. A lot of that investment never came back. I would rather we did not do the same thing again with AI.

In software, too, an “AI feature” is now bolted onto everything. AI should be the most useful thing there is for saving time and doing more with less, yet under that banner a strange amount of money seems to be wasted. Teams could pay a few thousand yen for a tool and try improving their own work — so why do so many companies pay millions to consultants who don’t really understand what is inside? I understand the panic of “if we don’t do it now, we’ll be too late,” but I suspect that panic is what makes the decisions worse.

However far AI goes, I think it should remain a tool that complements people. The heart of entertainment is the connection between people. The emotional resonance, the charm and the unpredictable fun that are particular to humans cannot be fully replaced by technology — and should not be.

Thinning language barriers, and cyber risk#

At the same time, I worry a lot about data privacy and cybersecurity as AI races ahead. As the technology improves, the risk that bad actors automate attacks on infrastructure is clearly rising.

Living in Japan, I feel strongly that the language barrier, which used to work as a kind of breakwater, is being worn away by AI. Many Japanese companies still seem unprepared for the sophisticated social-engineering techniques that overseas attackers use. Add AI to those attacks, as will surely happen more and more, and honestly, it is frightening.

Entertainment, with its many digital interactions and the amount of data it handles, may be especially exposed. Stronger security, better staff training, and real awareness of how AI-driven threats are evolving feel urgent. Even so, I can’t shake the worry that we may not keep pace.

Why I still think we should use AI#

In the end, I believe AI used carefully and responsibly can benefit entertainment enormously. Supporting people’s experiences, powering marketing, streamlining operations — this is no longer potential; it is already happening. To get the benefit, though, we need constant vigilance, robust security, and above all a stance that keeps human connection at the centre.

Put simply, I see LLMs like ChatGPT as technology for seeing patterns at scale. That is why I think they should be used first — especially through their APIs — to understand, analyse and automate the parts of entertainment where patterns are easy to see. Not AI as the star, but AI as infrastructure that makes human creativity and connection stronger: whether we manage that may be the dividing line for AI over the next few years.