Six Disruptive Ways Generative AI Could Evolve — and What That Means for Us

Six Disruptive Ways Generative AI Could Evolve — and What That Means for Us

Welcome to Leading Disruption, a weekly letter about disruptive leadership in a transforming world. Every week we’ll discover how the best leaders set strategy, build culture, and manage uncertainty all in service of driving disruptive, transformative growth. For more insights like these, join my private email list.

I’ve been following technology for 30 years. I’ve seen a lot of disruptions — from the internet and search engines to social media.

But generative AI just feels different.

It’s simultaneously fascinating and nerve-wracking. And it has a lot of implications for us as individuals and as a society.

Why is that? What opportunities are there? Are there any downsides? On Tuesday’s livestream, I explored why this is the most disruptive time I’ve ever experienced, what that means, and where we’re headed — and no, it’s not about mass unemployment or an uber AI taking over the world!

Why does this feel so disruptive?

Power shift create a sense of disruption. And AI is creating power shifts on multiple levels.

AI and specifically tools like ChatGPT are changing the ways we conduct ourselves, communicate with each other, do business, and exchange money.

It’s shifting the power balance. Suddenly, we have more capabilities than ever before. We have the power to do things that once felt impossible. We can write things, make pictures, and create code. (I’m coding for the first time in my life, which is mind-blowing to me!)

It’s generating unprecedented amounts of creativity. We can input information, and AI spits out something intelligible and well-formed.

It’s shortcutting the heavy lifting we’ve done in the past — but also challenging us to add more intelligence and creativity, to rise to a higher level. 

And all of that adds up to some exciting (and anxiety-inducing) opportunities. So:

Where does AI go from here?

AI is reshaping the way we do things. We’ve already seen impacts on jobs, mundane tasks, and content creation, among other things.

But what’s next? What will AI look like in the future? Here are six disruptive ways generative AI might evolve — and some of the implications we could see as this technology continues to develop.

1. Digital assistants

Just imagine having a truly intelligent copilot, a digital assistant, a chief of staff who learns your preferences over time, ingests information from anywhere (and everywhere!), and has access to all your apps and data. Would having a teammate or colleague who could find information and seamlessly deliver it to you transform your day-to-day? I think so — and this could be a reality within the next year. It's already here in early forms with tools like AutoGPT and BabyAGI. (I'll be talking about this more in my livestream next week, more info below.) BTW, AutoGPT was announced on March 30th and BabyAGI made its first debut on March 26th. It's happening fast, hair-on-fire fast!

2. New skills

With generative AI, skill doesn’t come from having knowledge anymore. It comes from asking really good questions and training AI to learn from you. Right now, the more you chat with these tools, the more confused they get. It’s hard for them to keep track of things over a long period of time. But a new companion that works alongside the AI, guiding it and feeding it information based on your preferences, can help you apply learnings and use them over and over again.

3. Enhanced software interfaces

I really dislike my CRM. I hate entering my contacts, tracking activities, updating them, setting reminders. I just want to get off a call or finish an email and know the information is inside my CRM. AI could provide a completely different way of using applications and designing software in the future so we no longer have to hunt and peck for information — we can just have AI handle it. 

4. Personalized education

Teachers can create personalized education for gifted students or those who want to go faster. Or they can use generative AI to develop lesson plans for students with learning disabilities. My son is dyslexic, and he loves using ChatGPT to help him structure his first drafts. It gets him over the first hurdle so he can move on to the more creative aspects of his assignments. Having a tool to fill in the executive functioning skill can be life-changing for students.

5. Faulty data

AI offers plenty of benefits, but one of the biggest downsides is the concentration of power in the hands of people who make decisions with faulty data. They could reproduce existing systems of oppression, and we might not even be aware of the damage to our information ecosystem. Or you may be using customer information that you don't have permission to in your AI models and may have to scrap your development efforts. Knowing what's in your data and your rights to use it is foundational.

6. Energy use

Generative AI uses a lot of energy, which has implications for our natural ecosystems, too. Having digital assistants requires a lot of computer power and electricity and other sources we might not even be aware of yet. 

One of the most disruptive things about generative AI is that it makes us question the fundamental ways we look at the world. It encourages us to reconsider how we make decisions and how we gather information. As disruptive leaders, it’s our job to create transparency and accountability, to use the power in an ethical way, and to establish principles of fairness, privacy, security, and inclusiveness so we’re making those decisions and using that data responsibly.

Next week, we’re continuing our exploration of generative AI but turning our attention to some newer tools, like AutoGPT and BabyAGI. I’m excited to dive deeper into the world of digital assistants, and I hope you’ll join me on Tuesday, April 25 at 9 am PT for another fun conversation!

Your Turn

Please share something you’ve done with generative AI — whether it’s with a Bing search, Bard, or ChatGPT. What have you been doing? What are you exploring? How are you integrating these tools into your daily life? I’d love to hear your experiences!

Panteli Tritchew

Educator, Communications and Entrepreneurial Leadership, Kwantlen Polytechnic University

1y

Thanks for another provocative thinking challenge, Charlene. Yes, AI is shifting the power balance and gifting us with more capabilities than ever before. And with more power and more capability comes more responsibility for those who embrace this latest iteration of digital evolution: responsibility *to* our society, our community, our employers/employees, and ultimately responsibility *for* our actions, our work, our content and our output. I appreciate that you end this piece with the caveat to use the power in an ethical way; that is, to keep an eye on our input. We don’t want to fly too close to the sun… 😉

William Tarpai

Achieving successful Sustainable Development Goals outcomes in the US and Globally

1y

Slow in responding as I have been exploring the Dutch cities and nature. One of the things that intrigued me visiting the International Criminal Court (ICC) was its efforts to get people involved with the Sustainable Development Goals challenge, especially SDG 16 as it relates to the court of last resort for international crimes against humanity - https://www.icc-cpi.int/get-involved/sdg16. Visits to icc. https://www.icc-cpi.int/contact

Like
Reply
Allen Borts

Applied Direct Services Corporation (ADSC)

1y

Hi Charlene I've now been working with the kinds of tools intensely since last November when ChatGPT first came out. And then more recently very intensely with Bing. I've done complex analysis and tested it. For instance on business processes in my company. I found the results very amazing. As you say it depends on how you ask questions. I've spent a career asking non-obvious questions in business. Often people try to second guess you and get things wrong. Intelligent machines seem very similar in that respect. Years ago I joked that if the machines become too smart I would go back to school and get a degree in machine psychology to deal with the same level of ambiguity as human beings. Today that is no joke. It's pretty real. I've been involved in one way or another with AI for decades. So, what's the difference in 2023? It's like you can walk into a.meeting in real time with the equivalent of 20 or more experts all at one time. That is if you personally have the depth and power to ask the right questions to channel all that. The problem is not the answers. The medium is the questions. This is why we all need more consciousness awareness to pick up such capabilities and use it for good

Daimean Graham

Daimean enterprise Close store

1y

God bless 😝 🙏 ❤️ 🌹 ☔ 🌈

Like
Reply

To view or add a comment, sign in

More articles by Charlene Li

Insights from the community

Others also viewed

Explore topics