If you are a Twitter or LinkedIn user you would think everyone in the world was embracing AI but the reality is this is an echo chamber and even with all the likes, retweets and comments the number of people who are using these tools to assist their day to day work is minor.
Polls and surveys don’t mean much on a small scale but I asked a random sample of thirty people I know outside of tech if they were aware of AI and how many are using it in their day job. The results were of thirty people asked, five people were aware (mainly ChatGPT) and two had used it. Of the two that used ChatGPT, one had used it to write a resignation letter for a family member and the other had used it for blog ideas for her marketing job.
We are truly in the hype cycle of AI but over the next nine months things are going to settle down and this technology shift will become accepted (we are running at that pace). This will come in small steps like AI-powered features being added to our everyday tools. We see this already in modern tooling such as Notion, Canva, Slack, and Webflow but the titans of enterprise Microsoft, SalesForce, Google, and AWS are all rapidly adding AI to their platforms. This is an unusual step in that enterprise is running at the same pace as silicon valley unicorns showing they understand what is at stake here.
Though like with all technology shifts, AI is no different there will be slow and fast pace switches, slow paces are the extra features we are seeing now:
Canva adds an AI background image remover.
Notion will summarise your meeting notes.
Webflow enables to you change text into a softer tone.
The fast pace switches are things that change our default behaviours:
Calling someone instead of using postal mail.
Buying a car via a website and entering your card details rather than meeting a guy in a car park with a bag of cash.
Or change how you have searched for information for the last 10 years via Google to now using ChatGPT which on the face of it “makes up its answers”.
These fast pace switches create leaps forward for humanity, the computer, the internet, the phone, 3g connectivity. The age of AI is the next leap but this time traditional enterprises have to act much faster than they did to previous technological shifts because…
You may roll your eyes at this next statement but I believe:
Within the next three years, a single individual equipped with AI will outperform a twenty person department inside of an enterprise.
The question enterprises need to ask themselves is "Do you have these AI-powered individuals inside your organisation and if not you need to get them or create them".
I am a firm believer that AI is going to change over 50% of jobs in the coming decade, it was always on the cards that AI would create this shift but the recent convergence of the following has accelerated this shift dramatically:
Web3 tech VC investments
The global recession and mass layoffs
Consumer-friendly AI tools
VCs from around the globe pumped into Web3 tech that hasn’t found mainstream consumer fit and those startups need to do something that returns and AI is it. So there has been a big shift in Web3/blockchain/crypto startups now becoming AI startups.
The global recession triggered largely by the recent pandemic has forced many companies to change from a growth hiring strategy to a "keep the share price up" hiring freeze and layoffs. As this settles down companies are still going to need to do more but with fewer people.
The AI tools of today are for the first time mainstream accessible, the over 100 million people globally using ChatGPT aren’t developers running command lines and adding Python libraries. These tools are giving domain experts the ability to do things they couldn’t dream of a year ago and all via an accessible UI and the programming code of natural language.
So what do you do if you are an enterprise?
Step 1 - Let people play
We don’t know what we don’t know and most of those in enterprises don’t know what AI tools can do to drive their output or remove mundane tasks. Identify tools with AI at their heart like ChatGPT and MidJourney and let people play.
Step 2 - Work with IT
There is a lot of fear in letting employees use products like ChatGPT, the concern is that corporate secrets may get shared and there is a lack of trust with big tech on what will happen with that data.
Bringing IT into the conversation and taking a “how can we embrace this” rather than a “what are our risks” approach means data privacy and security can still be at the forefront but doesn’t mean Tom in marketing or Sally in accountants can’t ask ChatGPT to summarise a report they are reading that is publicly accessible on the internet.
Step 3 - Share and Showcase
Showcase the interesting and amazing outputs your early adopters are doing with this technology. This process should already be in place in most enterprises as a part of the digital transformation programme.
However, you have to be accepting that the results may be uneasy, digital transformation has been the process mainly to upgrade and improve digital systems and the people have been brought along for the ride.
The shift to AI is going to show how much a single person can output with the assistance of AI technology compared to not just another person in the same role but a single person vs a whole department. You need to be open to this.
Step 4 - Incentives
The recent global layoffs of knowledge workers have created whispers that AI will just be brought in to replace these roles. Companies are already replacing customer service agents with automated AI systems so expanding this to other areas isn't that crazy and as the output of LLMs gets more trusting so will their use on the frontline.
Those that can captivate these technologies and tools are going to be in high demand, this may mean a move away from hourly remuneration is needed. How do you price one person utilising ten AI agents, do you give the “you should be glad you still have a job” line? This isn’t going to work this time, if you have a single person who can output what your whole marketing team could last year then that person, if not incentivised on output, isn’t going to be staying they are going to be set up on their own.
Step 5 - Acceptance
I have now accepted that business is an AI-based one, from simple things such as generating meeting notes to the complexities of drug discovery. The same way that only ten years ago we were advocating that everyone was a tech company, they just didn’t know it. Today every business is now an AI model and you better hope it is you and not your competitor.
And for those who read this and are sitting back thinking their regulated industry or supply chain is your defence, well it is but not for as long as it has been before because a guy called Amir in Egypt or Rosey in Denver who has twenty years of logistics knowledge is going to build an AI-powered shipping platform in three years on their own with hundred AI-based assistants and is going to become your supply chain partner.