March 20, 2023

How to Protect your Business from Generative AI Disruption

I’ll reluctantly admit that I did not immediately grasp the dramatic impact the launch of Chat-GPT would mean for B2B Tech companies. I’ve been writing, presenting, innovating, and podcasting about AI and its impact for years, and generative AI was not new, so initially I viewed it as just the next step.

But something I read in the NY Times last week woke me up to the new reality. Open AI did a live demo in which it fed GPT-4 (the newest evolution of the model behind Chat GPT) a picture of a diagram of a website, and GPT-4 coded and built the website, completely functional.

The simplest way I can think of to explain why LLM (Large Language Model)-based generative AI tools like GPT-4 represent a dramatic paradigm shift in business is that AI has historically been on the backend. It has made recommendations, optimizations, and determinations, all behind the scenes.

Suddenly, AI is now taking a starring role on the front end. It’s offering a new way for consumers and workers to directly interact with technology, and it creates the results from technology that consumers and workers directly experience and use. That’s a massive change.

So anyone that thinks of generative AI as a chatbot or a search engine has it all wrong. That’s like thinking of TV as visual radio.

Any smart business leader should already know the urgency of this paradigm shift, and should be actively trying to figure out specifically what to do. This article will provide a framework for how you as a business leader need to prepare, so your company will be a winner, and not a casualty in the explosion of generative AI.

At a high level, you need to think through three categories: People, Defense, and Offense. Let’s take a look at each one:

The impacts of generative AI on people can be categorized into training and staffing. First and foremost, you need a team that intuitively understands how AI functions, and internalizes how its skills are complementary to those of humans. Too much of AI-oriented education is technical, i.e. what’s a Large Language Model and how does it work?

That’s not the right mindset.

What your company needs is not about technology or data or math or coding, but rather it’s about intuition. I’ve had a lot of success training people for years on AI by using dogs as an analogy. How do dogs learn, and what are the limits of their abilities? How are those similar or different to today’s AI models (which refers to “narrow AI”, not the long term promise of artificial general intelligence, i.e. “AGI”)? The bottom line is simple: You need the right training in place for your team.

As for staffing, there are a few key questions to answer:
– Which specific roles can generative AI functionally replace in the near term?
– For those roles, does that mean downsizing (AI as replacement), or does that mean expansion and growth in areas where staffing up is difficult (AI as complement)?
– Which new roles are required to ensure your team can best function as a co-pilot for generative AI? What does co-piloting mean for your use cases? What sorts of inputs are required for generative AI to be effective? What sort of reviews of generative AI outputs are necessary to minimize bias and maximize trust?
– Can existing employees be trained to take these roles? Or are new external people needed? What do job descriptions look like, specifically?

Smart business leaders need to immediately and realistically assess the direct threat generative AI presents to their existing business. Why? Because nearly every business will likely experience some significant degree of disruption. In general, hardware businesses are better situated than software businesses, because they operate more on the backend. Virtually every software business is right in the line of fire.

Generative AI will change how customers interact with technology, representing a change from doing to getting. For instance, imagine when you call a support phone number for technology today. What do you tell them? Likely you tell them what you’re trying to do, and where the process is failing. That will change. In the future, we’ll instead focus on what we’re trying to get from technology. That’s an entirely new mindset.

Think about how customers currently interact with your tech. How could – how will! – generative AI change the inputs? To what extent does your business need to rethink the entire customer experience to adapt to the users of the all-too-near future? What will users expect?

Here’s a scary thought: Could a new entrant (or old rival) avoid a lot of the baggage of a legacy UI and build on top of generative AI to create what users expect more quickly and efficiently?

But responding to the advent of generative AI is not only about fending off disaster. It’s also about taking advantage of new, potentially once-in-a-generation possibilities: recognizing and executing emerging opportunities created by generative AI faster and better than others. Companies that thrive in a generative AI world are likely to have the following assets, and to use them to disrupt the status quo in areas outside their core businesses.
Unique data at scale: As you’ve likely heard, one of the limitations of Chat-GPT to date is its proclivity to extend beyond facts to generate its own plausible-sounding falsities. That’s an artifact of poor training data. Right now, Chat-GPT is functioning as a jack-of-all-trades, and AI is simply not great at that. As generative AI matures, and as companies build these LLMs on top of custom datasets for narrow use cases, it’ll develop into niches. That’s when the results will become strong and reliable and thus scalable. You need to immediately evaluate the following: Is your company in a position to win in this environment? Does it have a scalable and reliable source of unique and valuable data? Is that dataset thorough enough with a diverse enough set of sources and subsets to power AI that minimizes bias and creates reliable results? If you don’t have any or all of this, where and how – and how quickly – can you get it?
Reach: Generative AI is now officially a race, with early mover advantage being important. But success isn’t just about ideation or even product innovation. It’s also about reach, i.e. how widely can you get customers to try your new experience, get hooked on it, and embed it into their everyday lives and processes? This comes down to the size of your customer base now, the power of your integrations and partnerships, and your voice in the marketplace. Now is the time to fully leverage all of these.

There is much more to know about how businesses should respond to the emergence of generative AI. But hopefully this brief discussion functions as an initial framework to help smart business leaders like you move from thought to action in the race to win in the generative AI world.

We are all running out of time to be early. If you’re looking for customized guidance in implementing this framework, and creating a clear plan to leverage generative AI, please reach out to me for a free initial consultation.