AI in 2025 - a year of change

We saw plenty of progress in AI in 2024 though most of the headlines were grabbed by the big players such as OpenAI and NVIDIA in terms of progress in larger, more capable LLMs and faster, more powerful chips. In terms of practical examples for AI, or specifically generative AI, too many use cases are focused on mundane, low-brow content generation. There’s a lot more to AI but adoption has remained slow even in the US and advanced economies. Go on to most websites and you still must navigate an FAQ to glean information that may or may not be elsewhere on the website. Yet, replacing an FAQ with an AI chatbot takes less than a day to implement (for most websites) and has minimal associated risk.

The inertia of some businesses does not augur well for their future. Entire industries are poised for disruption, and there will be clear winners and losers. We will have Kodak-like extinctions and new global players, currently being incubated in a bedroom or garage somewhere (hopefully belonging to someone you are related to!), will emerge.

We see the changes in our own industry - there are clear signs that coding is no longer the exclusive domain of software developers, nor AI/ML model building the preserve of data scientists. AI-powered tools increasingly handle routine programming tasks, leading to a perceptible slowdown in tech recruitment. The same seems to be happening for content producers, as AI generates articles, images, and video/audio at minimal cost. In fact, the cost of generating content is dropping so sharply that AI threatens a tsunami of cheap content that most people won’t read or watch. There is an upper limit to the amount of content the world can consume, and cheap content will just wither on the vine. Quality and originality will be crucial because the average consumer’s attention is already stretched.

Standing still is not an option. Customer service, for example, is likely to undergo a massive shift within the next year, delivering improved satisfaction at a much lower cost – something that is long overdue. We, the consumer, have suffered years of poor customer service, AI might finally change that though the pace of change amongst many big brands is glacially slow.

Below are a few recommendations if you haven’t already begun your AI journey:

1. Create a clear AI strategy that is aligned with your core KPIs.
2. Start small and expect mistakes
3. Focus on manageable use cases that don’t risk the farm
4. Keep learning, things are changing quickly.
5. Stay on top of changes - competition will intensify and tech is evolving at a furious rate; be prepared to adapt quickly.
6. Stay ahead of regulation - the EU AI Act, like GDPR, will have global implications.

?

Imagine you were a manufacturing company in New York in 1882 when Thomas Edison opened the first commercial electricity company – would you (a) ignore it, since business has been going fine or (b) figure out how you can adapt this new technology to grow your business? For anyone with doubts as to the transformative potential of AI you might want to consider AI in the same light, pardon the pun, as the Edison Electric Light Station. ?It’s here to stay and will change the world.

?