Thinking sensibly about AI for your business
Navigating the hype and the scepticism around generative AI
People in tech seem to talk about ‘doomers’ and ‘accelerationists’ primarily in the context of AI safety discussions - aka whether or not they view AI as humanity’s downfall or saviour. But after some conversations in the past few days, it strikes me that there’s resonance in the more mundane setting of business too.
For a flash in the past year there was frothy excitement (and some trepidation) about generative AI’s potential to Change Everything. But after a few well publicised debacles (e.g., Air Canada’s chatbot fiasco, the lawyer who didn’t check ChatGPT’s citations, Uber’s problems with ID verification) the mood has started to shift. Gartner seems to have called it right in August 2023, positioning generative AI at the peak of inflated expectations and about to drop off the cliff into the trough of disillusionment.
So maybe not the most opportune moment to be starting an AI consultancy(!) But to think this misses the forest for the trees. AI - or at least generative AI - has been ridiculously over-hyped in its short term impact, and it’s helpful to bring it back to business reality. A healthy dose of scepticism is always useful when making decisions to embrace technology, to ensure they’re rooted in a purpose beyond just keeping up with the Joneses.
That said, there are many real opportunities today to benefit from generative AI, with more on the horizon. So even if you’re not ready to embrace such tools now, you still shouldn’t ignore them. Look closely at that hype curve, and at how long they expect generative AI to take to get to the ‘plateau of productivity’ - just 2 to 5 years! In the same way that generative AI stormed to peak hype in record time, despite speculation there are no signs yet that it won’t continue its barnstorming progress in capabilities. And even if advancements were to slow, there is still plenty of value to be tapped in existing tools as people figure out better ways to use them.
So if you’re still on the sidelines, wondering about whether or how you should think about AI in the context of your business, here’s a few provocations to get you started.
» Know your lane: do you want to work in AI? Or with AI?
(kudos to my friend Nigel for coining this phrase that sums it up so succinctly). The right answer for the vast majority of companies will be the latter. Too many businesses have gotten this muddled, thinking they need to build their own language models and tools. That’s like deciding to expand your company headquarters, and opting to make your own bricks, smelt your own steel, and DIY assemble it from scratch. Nowadays, enterprise suppliers of AI tools offer options that will get you better outcomes, faster, and likely cheaper and more reliably too.
» AI gives you a different drum, but it’s mostly still the same business beat
Hold tight to traditional business principles when evaluating opportunities for using AI. Embracing AI should solve a business problem, not be a goal in itself. Yes, AI may let you do things that were not economic before. Yes, AI may introduce challenges of a different nature and risk profile. And yes, AI is likely to be disruptive in the same way other technology shifts have been, necessitating a rethink on operational processes, how tasks are parcelled up and what this means for skill sets and staffing. But, for most brands, the fundamental underpinnings of your business are likely to remain unchanged. So use those as a rudder to navigate the myriad of opportunities, prioritising those which make the most sense for your business in its current context.
» If you wait, do so with intent — not in the dark
Inaction is not neutral; it’s an active decision to stand still. For some businesses, waiting might be the right decision, for now. But make sure it’s not just a default ‘head-in-sand’ move or that you’re frozen in place by indecision. You should be able to articulate why exploring AI opportunities isn’t right for your business at this moment; and identify the signposts that will signal when it's time to reconsider and ramp up. Ask yourself: if not now, why not – and when?
» Don’t be naive about generative AI’s creative capabilities
AI is statistical algorithms, sure. But don’t let the how distract you - whilst generative AI may not be truly creative in the human sense, it can still produce highly creative, innovative output – much faster than you or I. Arguments that AI output will be ‘too perfect’ or shiny to seem authentically human assumes that it cannot mimic imperfection – but it can (remember Google Duplex?). It is also wishful thinking to dismiss AI’s creative potential by setting the evaluation bar to the most extreme: AI acting alone to produce the final output. Maybe AI will reach that pinnacle, maybe not, but in a way it’s moot because generative AI is already such a powerful tool when harnessed by creative people. And as for cries that AI creative is inauthentic? 19th century painters felt the same about photography.
» Look ahead: Today’s AI is the worst you’ll ever have
Yes, language models hallucinate, but techniques are emerging to improve control and make it easier to spot errors. People are also getting better at figuring out operationally how to deal with such challenges. From a societal perspective, we’re just at the beginning of the AI era – it’s like the Internet before the web. Few could imagine then how the Internet would transform daily life for billions. In 10-20 years we’ll say the same about generative AI. And just like with the Internet, the sooner you dive in, the better poised you’ll be to embrace the transformational advances in AI intelligence just around the corner.
So often, luck in business is about being in the right place at the right time. Sometimes you can create your own luck; sometimes it’s about being open to the possibilities that arise. In both cases though, you’ll achieve so much more if you’re poised and ready, having already built the muscles to take advantage. Engaging with AI now, even cautiously, ensures your business won't be left behind as the technology matures. It's not about rushing in blindly, but about making informed, strategic decisions that align with your business goals, and build your capabilities so you’re ready for the lucky moments to come.