Journalists love stories about AI. It allows them to come up with titles like “machines use imagination to solve problems” and “neural network invents language”. An alien reading those stories would think that humans have finally cracked it. We must be in some golden age of computing. That same alien would be shocked to find out that at any given office on any given day the problems are much more along the lines of “my CMS is driving me crazy; where is the image I uploaded?”, “I can’t download my emails again” and “I wish there was a CRM that actually worked!”. How come AI is not solving all these problems?
For small or medium-sized companies (and a lot of very large companies!) what does all this hubbub about AI mean? Is it just journalists hyping something to get more clicks? Do you need to care?
The short answer is: “Yes, you should care”. More importantly, you should prepare. AI technologies have reached a tipping point. They are going to change things. Companies that are prepared and can exploit the new opportunities will thrive while the rest will struggle.
To understand why AI is going to change things you need to understand what the real AI revolution is without the hype and the references to Terminator. Ready? Here goes:
AI is about delegating decision-making to software. Software is becoming better at decision-making because machine learning tools have become better at predicting outcomes and easier than ever to use. In limited, well-defined, contexts (i.e. a lot of work situations) delegation of decision-making to such software can outperform humans.
That’s it. AI lets you get more done, better. Like any good machine should.
With that in place what should you be doing today to prepare and take advantage of AI technologies. Here are 3 simple things any company can do:
AI thrives on clear instructions and lots of data. As a company you need to explicitly understand what it is that you do and what data you use to decide what to do next. By explicitly, I mean written down as clearly as possible. For example:
“To decide what action to take in step X of process Y we need to know the value of Z, which we compare to K and then decide based on whether K is bigger or smaller than Z.”
Congratulations! Having reached such a level of clarity you can delegate that to a machine.
Sometimes we don’t really know what decision-making happened that led to a desired outcome. Luckily, we can throw data at algorithms and uncover those hidden patterns.
Training neural networks to solve problems may sound like a far-fetched idea in your organisation today. That day will come though. When it does the first question Joanne, the machine learning consultant, will ask is: “Where can I get data from the past five years of process Y”. Joanne will be very happy if you can point her to properly stored and annotated data and Joanne’s bill will be significantly lower as a result.
Invest in understanding you data flows and storing data in ways that it can be effectively reused in the not-so-distant future.
No, I don’t mean expensive PhDs. I mean start using the wealth of tools that are coming from companies exploiting AI to solve problems.
Need a quick logo for a small marketing campaign? Try out Logojoy. Need to edit some text? Try Grammarly. Need to improve support? Use AI to create an FAQ or create a chatbot. Scheduling meetings drives you crazy? Hire a personal assistant. There are hundreds of solutions out there that will automate tasks for you. Start getting used to the idea that a lot more of what we used to do can be delegated to machines. Look for those solutions. Half the battle is understanding that you can look for such tools. This will help you think more clearly about what goes on within your own company that could be automated as well.
There you have it. Understand your process and properly manage your data. Hire AI tools and get used to the idea that you can delegate far more tasks to machines than ever before. This way you develop a culture and environment that will be ready to exploit AI solutions tailormade to your organisation.
Ignore the AI hype but make sure you don’t ignore the fact that a big change powered by AI is happening.