Some of the most impactful conversations happen in the most unassuming of places.
This article was sparked by a casual chat over a drink with a friend of mine, who happens to run sales and marketing for two of New Zealand's most trusted brands.
As we talked about the potential of AI, it struck me how important it is to articulate just how rapidly AI products and solutions are evolving. I don’t think the layperson is fully aware of quite how fast AI technology is developing. The pace is astronomical.
Also, as the conversation progressed, I recognised that the yardsticks we have traditionally used to assess software solutions will need a substantial rethink as we navigate this new world of AI-first technology…
What's under the hood of most off-the-shelf AI solutions?
Right now, AI startups are proliferating at an unprecedented rate. Every day there are dozens of new AI tools and companies surfacing. However, if you peel back the glossy interface, many of these so-called "innovators" rely heavily on third-party AI engines.
Of the current tools that generate written content - the majority lean on OpenAI's technology. Similarly, today most AI applications that generate images are built atop Stable Diffusion technology.
I see a significant conundrum ahead for these off-the-shelf AI products, building in this way…
They lack a solid moat, or a competitive advantage, that protects their business.
Why? Because APIs, the backbone of these tools that connect the underlying AI tech (ag Open AI’s or Stability AI’s models), are readily accessible.
Any budding entrepreneur can build a similar tool using the same APIs; leading to an inevitable race to the bottom with pricing. It's hard to outpace the competition when everyone's using the same wind.
Then, when you layer in the ever-looming competition from tech titans like Google, Microsoft, OpenAI, Apple, Amazon, and Meta, the ground that off-the-shelf AI tools are building on looks even less stable.
The AI behemoths have pledged hundreds of billions collectively for AI development, arming them with the resources to build and market tools much faster than most startups.
As Ken Smythe, founder and CEO of Next Round Capital Partners, anticipates, this potent combination of factors could lead to a startling 85% of AI startups falling out of business within three years.
It’s the old “are you building a start-up, or just a feature of a bigger platform’s product” conundrum that traditional software start-ups have faced. Just transplanted to the generative AI era.
Turning the problem into the solution
Thankfully, the problem is also part of the solution.
If anyone can construct a tool on top of the APIs from the leading generative AI providers and compete with AI startups, so can individual businesses - like yours.
I’m talking about the ease with which any business can now build bespoke AI solutions; designed for their own companies and the unique challenges and opportunities within those companies.
Until now, building custom software was a luxury reserved for enterprise sized companies. Those with deep pockets and a plethora of in-house development resources.
We’re watching that era die, before our eyes.
With the emergence of sophisticated large language models like GPT-4, the financial and human resource burden of developing and maintaining bespoke software and apps (powered by generative AI) is plummeting.
In essence, the work previously shouldered by custom code is now managed by these foundational AI models, which is significantly less complex to build with than traditional coding.
Imagine you're running a customer service department. Instead of investing heavily in training human staff or developing a custom chatbot solution using traditional software solutions, you can now utilise GPT-4 to handle a bulk of your customer queries in real-time, 24/7, and in multiple languages.
This significantly reduces your overheads (no need to custom build a chatbot using traditional coding processes) and scales up your capability (the heavy lifting is being shouldered by GPT-4’s advanced AI) compared to how the same solution would have been built 10, 5 or even 2 years ago.
This shift represents an industry sea-change, opening up avenues that were once deemed inaccessible due to cost or technical complexities.
So which is it; off-the-shelf or custom built AI?
The truth is that from now on, business leaders will likely be deploying a mixture of both.
As such, I see an era of tremendous opportunity unfolding for businesses.
As NZ business leaders, we will soon be spoiled for choice with a wave of off-the-shelf AI-powered software. These innovative solutions are poised to outperform anything we could have envisioned even a few years ago, providing the tools to revolutionise our business models.
At the same time, we are about to become the first generation of business leaders to witness and benefit from a sharp decline in the cost and complexity of building and maintaining bespoke, AI-powered software and apps.
Some use cases will lend themselves better to off-the-shelf AI solutions.
But I believe more use cases will be better suited to bespoke AI solutions. These will be rapid and cheap to build, and custom designed for the individual use cases of the individual business that’s building them. These bespoke Ai solutions will be plentiful, spread throughout any given business’s departments.
And they will be extremely powerful.
This coming shift towards attainable bespoke software will enhance our businesses in ways we haven’t yet imagined, should we take the opportunities.
Conclusion
To circle back to that casual chat over a drink I alluded to in my introduction...
I lacked the articulation in the moment to make my points clearly to my friend about how I see the approaching wave of attainable, custom built AI software for all businesses.
But it forced me to go away and do more thinking, so that I could have a second go at articulating myself properly on the topic, here.
To me, it’s clear that the rapidly evolving landscape of bespoke AI solutions in business is going to necessitate a paradigm shift in how we assess and adopt technology. It will be very different from how we assessed traditional software solutions for the past few decades.
The dawn of AI-driven, custom business solutions opens up a world of new opportunities for SME businesses.
As we sit on the brink of this AI revolution, I urge you to embrace the disruption, adapt to the changing landscape, and let's use the powerful bespoke AI tools coming our way, to shape the future of our businesses in Aotearoa.
Got something to add? Chime in below...