5-Steps To Fast AI Transformation
5-Steps To Fast AI Transformation
A guide for businesses and corporate teams to fight back the approaching competitive tsunami
If you’re a big or small business founded before 2024, or a team-leader inside a company that was, and you’re not sure how best to bring AI into your organization, you’ve come to the right place.
The first thing to be aware of is that it’s ok to have no clue. Nobody does really. Business leaders can’t possibly know how to apply the countless, exponentially evolving and multiplying AI tools and models to the myriad roles you have at your company. Not to mention that everyone that works for you probably has some vague sense of urgency and preference and it’s hard to know who to listen to.
Facts:
There are no experts. There is no one you can hire to fix this. There are too many AI tools for any one person to expertly track, internalize and lead systemic solutions for your business. At best we are all enthusiastic users. I see so many companies randomly purchasing AI tools at corporate rates because, frankly they don’t know what else to do right now.
AI tools are advancing and changing too rapidly to follow traditional software provisioning. Tool A might be regarded as the “best” today, but tomorrow it’s tool B. And I don’t mean “tomorrow” in the abstract future sense, I mean literally tomorrow. No sooner have you paid the annual fee to cover your staff, then its primacy has been upended. ...for a week, and then it’s tool C.
And finally yes, you’re rightly worried that some 13-year-old kid, named Spider, will develop an AI-based business in his bedroom that viably competes with and potentially kills your business. Access to AI has flattened countless barriers that in the past allowed your unique knowledge and expertise to dominate a market, and your company to grow and defend itself to date. The triple moats of knowledge, skill and scale can all be largely breached today by one 13-year-old kid named Moonhead - and AI. Suddenly, that kid has your business’ knowledge, skill, and development scale.
This last point is most important of all, because where last week you maybe had 3-7 competitors, all of which you knew well having researched their strengths and weaknesses, their market positioning and your differentiation, today you are facing an explosion of potential competition. Thousands of them. Encroaching on your market from a thousand unpredictable positions.
Your competitive landscape has gone from a few big rocks that you could see, aim and fire at, to a particle cloud of competition. And taking aim at a particle cloud is like shooting a gun into a swarm of bees. You might hit a couple, but the swarm will be fine. And many of them will still be optimized against you.
Unfortunately the number of ways AI will be employed by these thousands of uncoordinated individuals who would each like to compete with aspects of your business will be impossible to identify, track and defend against as a top-down organization.
It won’t work.
Individual people can be smart and quick.
But God love companies, even small ones, are big, dumb, lumbering animals; the product of slow and steady time frames for innovation and development. In companies, decisions take time requiring meetings on Wednesday, cross-division consensus, annual budgets, and contract reviews. It even took an extra fiscal quarter when that one email got stuck in Steve’s outbox for a week so Caroline couldn’t print it in time for the board meeting in August.
Whereas right now some kid named Zeus, using AI, can stumble upon an idea and deploy a functioning solution in minutes.
You can’t stop this pending tsunami of competition. It’s coming. And whatever limitation AI might have today that you’ve been reassuring yourself with, it will vanish before your memo gets to Dave in accounting.
Software companies are obviously most at risk for the moment. That’s the war’s front line.
Companies whose products or services are all or partly real-world-based will have a little more cushion, but any aspect of their operations or processes that can be automated by AI or any proprietary software systems that power them will be at risk of being done faster and better in a thousand ways.
So what do we do?
You compete with a particle cloud by becoming a particle system.
Fighting the cloud with a system
A company of any significant size is a potentially formidable force if it frees its agents in pursuit of solutions.
Sufficiently empowered, your employees can save your business.
Many months ago, my team and I made a key change in our work week. It was initially just a test, a conceptual trial. And as with all tests it might’ve been a huge failure. On the contrary it has proven to be wildly successful. With very little adjustment, our team has since settled into a new way of working that’s resulted in the rapid generation of countless new talents, ideas and AI solutions.
Whether you are a business, or a team inside one, I strongly recommend you follow these 5 steps:
1) Redirect 10% of your employee’s workweeks to AI
Going forward direct your teams to spend 10% of their work week, free of their daily tasks, exploring AI.
At first your team won’t believe you.
My team does this on the first half of every Friday. I chose the first half of Friday, as opposed to the second half, because I want their best, fresh morning energy and attention. Not when they have a foot out the door for the weekend. The team looks forward to this weekly event.
Yes, this change implies a 10% reduction in productivity. Suck it up, own it. Let them know they have your permission. Let them know this is their job. Mission critical. For some of you a 10% reduction in productivity will sound untenable at first. But the benefits that are gained on the back-end of this process through the resulting solutions (AI-powered solutions and tools) will more than make up for your 10% productivity loss. It will come back to you in multiples.
2) Give your teams freedom
Empower those teams to explore AI during that 10% in any way they wish. Let them follow their creative muse, whether it seems work-related or not. For some that will mean they will search for specific solutions to overcome problems they face in their roles, or they’ll explore whole new opportunities for the business. You will find most of your staff naturally leans this way; people want to be useful. However, you will also find a number of employees who explore AI in a completely different way. You might be tempted to characterize those people’s time as “off target, a waste of time.” A finance person might create music, someone in HR might create a totally unrelated app, others might create artwork or videos. Don’t imagine this as any kind of failure.
You are empowering your teams to familiarize themselves with AI. Everyone has a different way into discovering a complicated, wholly new domain like this. The more your teams enjoy the process of discovery, the more they will learn, and the more varied and interesting your resulting ecosystem of innovation will become. At the very least, in time, this experience will translate for each of them into a foundation of understanding that benefits their future AI problem solving. You are giving them understanding.
What makes the evil cloud of competitors such a threat in part is the impossibly varied range of discoveries that could upend your business. It could come from anywhere. You simply cannot predict specifically which doors will lead a person to innovate a specific competitive solution. With AI, brilliant ideas will fall from places you can’t imagine. So don’t micromanage this process. Let go. Allow your teams to indulge themselves, just like the evil particle cloud does. They may make discoveries outside their job description, but it’s possible this will trigger ideas for someone else who does work in that field, or perhaps this is how you will discover a latent talent or a miss-assigned genius on your team.
Discovery, learning, and individual innovation is incredibly valuable, but only half of this process. Now you must turn your own particle cloud into a particle system.
This is the part of the process that some 13-year-old in his bedroom named Durk-Durk can’t do with AI alone:
3) Create a schedule of sharing sessions
My team meets every second Monday for one hour. We initially met every Monday - but this was too fast for most of us to internalize our discoveries and projects and present them back to the team. Every second week was the goldilocks timeframe for us. This sharing session is a chance for participants to present the explorations and discoveries they’ve made over the previous two weeks. To share the tools they used, and describe any unique strengths, difficulties and limitations in those tools that they encountered.
This is key because now we are making the rest of our team even smarter.
Everyone is learning the lessons of the others. Very quickly you will notice valid ideas or new methods appear. Tribal knowledge will take root. Through ensuing discussion, the best ideas will become strengthened by other participants, and new ideas will often find unexpected tactical or strategic purpose and supporters elsewhere in the company.
4) Outreach to other teams and executives
Depending on the size and structure of your organization, for example if your company is very large, you may wish to establish a selection process where a subset of the new ideas are shared broadly between divisions. At the very least we have found an evangelist of each big idea needs to reach out to other teams or executives to pitch the new solutions.
Trust me, when other members of the company see a smart new solution actually functioning, it will drop jaws and get attention.
5) House-keeping: Throughout, use introductory accounts and anonymous information
There might be better ways, but this is what we do:
Instruct your teams to take advantage of free, introductory accounts, virtually every new AI tool has them. If they don’t - do not let staff be tempted to pay a lower monthly cost for a year. One month is all you need to evaluate a platform. Let your team expense it. Most likely you will not want to go forward with this particular tool past one month anyway. These tools are changing way too fast to commit to a year’s subscription, unproven.
It’s your call of course but we always anonymize our information and data for privacy issues and proprietary concerns when evaluating new external tools. It’s tempting not to, but considering the vast range of tools out there already it’s impossible to know entirely how your information is being used. I’m sure there are countless more security concerns, and they will often be a reason not to explore and innovate. You need to determine what’s an appropriate balance for your team. Ours leans heavily toward innovation and learning and occasional associated risk.
At such time that a given tool proves itself or becomes a key part of some new idea, work with your organization’s IT teams to secure a wider subscription and secured corporate tier. At this point you can blow out ideas with company data.
If you are a team inside a business, and you have sufficient autonomy to enact this 5-step process, your team will rapidly become regarded as an innovation center for the company.
If you are a company leader, your business will stand a chance of beating back the randomized firehose of AI-powered 13-year-old competitors you’re about to face.
This process of atomized exploration and individual ideation inside your company, reflects the very same particle-cloud dynamic of AI innovation and development we have begun to see in the outside world.
Only better, because you have resources and the ability to coordinate your innovators to make something better and faster.