The AI Tsunami
In recent years, Artificial Intelligence (AI) has been heralded as a revolutionary technology - it will create breakthroughs in science, streamline industry and automate transportation.
The value to date has been in the creation of those models. Large technology companies have invested heavily in developing algorithms that can predict what you will type, automatically categorise images and flag misuse of social media.
But this value is changing. The advent of Large Language Models (LLMs) and NLP are indications that the models themselves are tools to build on top of, not the end result.
Models such as GPT-3 or Stable Diffusion are akin to AWS. They are AI infrastructure to be built on. They are a business in their own right, but the present the foundation to create huge numbers of business on top of them.
The Data is the Product
This leads to a clearly indication of just how critical data is in creating the next generation of systems.
The software industry had been build on developing tools to replicate, improve or automate processes. It is all about processes.
Take Hubspot CRM for example. It is data (about your customers, their companies, the interactions you have with them) with a process layered on top - create an account, record an email, create a sales opportunity, or support ticket. The software is a way to interact with the data.
The AI emergence outlined above will change this. Software will be commoditised, and instead of spending months building an app to recreate a process, you may spend much less time on this, and focus your attention on understanding how your data could be valuable (within and without your organisation) and how you could leverage AI to accelerate delivery of this value from the data.
If this is to be the new reality, then you need to manage data. You need to focus on data. Ultimately your product is the data it collects and creates, and this is what defines the value to your organisation.
Interfaces: connecting data to humans (and data)
So, to recap, we have AI which is now available to point at data, and return answers, results, or predictions. We therefore need to optimise (and for many, change) the way we organise data in our business. But this needs to have some end point. And this is the interface.
I use the term interface, because this could be a range of things. Yes, it may be a mobile app or a website. It may be a smart speaker. It may be a taptic feedback in a steering wheel. Or indeed the interface may not involve humans at all and interface directly with another machine or AI system.
Design can create empires - Apple is the most obvious example of this. You may have all the best data and best AI in the world, but if its painful to use then it’s unlikely to be a successful product. Only by combining all three - data, AI and interface design will you create a scalable, impactful solution.