Thank you for being part of ICMA 2024
(if you’re here, you’re already in the 1%)
KEY TAKEAWAYS (an AI-Output with Hallucinations)
Key Takeaways from the Speech:
The Invisible Nature of Trust and Its Evolution with AI:
Trust is a fundamental yet often unseen component of daily interactions.
The rise of AI challenges traditional notions of trust, prompting questions about how trust will evolve as AI becomes more integrated into society.
With predictions that up to 90% of internet content could be AI-generated by 2026, distinguishing between human and AI output becomes increasingly difficult.
AI as a Democratizing Force in Work and Action:
AI is not just democratizing knowledge but also the execution of tasks, effectively transforming how work is done.
Individuals now have access to tools that can perform complex tasks, leveling the playing field and enabling anyone to create and innovate regardless of technical skill.
Transformative Applications of AI in the Public Sector:
AI has significant potential in areas like data analytics, customer service, and process automation within government services.
Examples include:
Data Analytics: Utilizing AI to analyze large datasets quickly, revealing insights such as parking violation hotspots.
Customer Service: Creating AI avatars to interact with the public in multiple languages, improving accessibility and efficiency.
Rapid Development: Generating functional prototypes or solutions (e.g., websites, apps) through AI, significantly reducing development time and cost.
The Emergence of AI Agents and the Reasoning Era:
Advanced AI models are now capable of reasoning and problem-solving, not just predicting text.
This advancement leads to AI agents that can perform complex, multi-step tasks autonomously, signaling a shift in how work is approached.
Redefining Leadership and Skills in the AI Age:
Leaders must adapt by leveraging AI or risk becoming irrelevant.
The most valuable skills are shifting from technical knowledge to creativity, innovation, and the ability to challenge the status quo.
Education and skill development should focus on fostering these abilities rather than solely on traditional programming skills.
Infinite Leverage Through Technology:
AI and other technologies provide unprecedented leverage, allowing individuals and organizations to do more with less.
This shift enables the reduction of bureaucracy and empowers employees at all levels to become creators and innovators.
Personal Responsibility in Adapting to AI:
Individuals should proactively engage with AI technologies rather than wait for organizational mandates.
Experimentation and hands-on experience with AI tools are crucial for personal and professional growth.
The Dual Nature of AI Advancement:
While AI offers immense opportunities, it also poses risks such as deepfakes, scams, and the erosion of trust.
Society must navigate the challenges of misinformation and develop strategies to maintain trust in an AI-saturated environment.
The Need for Self-Disruption and Challenging Status:
To stay relevant, individuals and organizations must be willing to disrupt themselves and reassess established norms.
Letting go of traditional status and embracing new ways of working with AI is essential for progress.
Embracing Fear as a Catalyst for Innovation:
Fear of new technology is natural but should be harnessed to drive learning and adaptation.
Being bold and willing to take risks leads to personal empowerment and organizational advancement.
The Importance of Critical Thinking and Skepticism:
In an age where AI can generate indistinguishable content, critical thinking becomes paramount.
Trust should be placed in verifiable work and outcomes rather than appearances or positions of authority.
ai TOOLS THAT SHAWN USES:
Perplexity - Instead of Googling (I like it because it gives you the sources)
Vercel - Building websites and prototypes
Claude 3.5 Sonnet - Feed data into Claude multi-model trend analysis, code, prototypes and gap identification
HeyGen - AI Video Avatar
Midjourney - Pictures and graphics (Much better than DALL-E)
Runway ML - Text-to-Video Generator
Clay - AI Outbound Sales | Generating lists of emails and sending cold outreach
Figjam - Whiteboarding Ideas | Brainstorming | Organizing
ThingSpeak- Collect and analyze IoT sensor data
SerpApi to analyze Google's "People Also Ask" to uncover unmet needs
Iris.ai or Consensus.app - for Al-powered research exploration
CLEANED TRANSCRIPT (an AI-Output with Hallucinations)
Thank you so much. I'm so excited to be here today. It’s interesting—over the last five minutes, I made about a thousand trust decisions in my mind. I had to trust that Lon introduced me correctly. I had to trust that you would clap. I had to trust that you’d actually be here, even though there’s a Steelers game happening just five minutes away. I had to trust that this clicker would work, that the screens would function, and that the AV team knew what they were doing.
I made a thousand trust decisions in the last five minutes. And that’s the beautiful thing about trust—it’s invisible. But for the first time ever, I’m trying to make trust a little more visible. I’m going to ask you a bunch of questions, and I want you to clap or cheer if you trust these individuals, products, or companies.
Let’s start with the first one. Clap or cheer if you trust Apple. Alright, clap or cheer if you trust this person. Okay, the gentleman in the front says, “I do not trust Oprah.”
Do you trust an AI driverless car that’s rolling past you? Clap or cheer. Do you trust this person? So, you trust the AI, you trust the cars it’s making—just not the person. Do you trust Microsoft Excel? Of course, it’s the most trustworthy of all time. And do you trust ChatGPT?
My observation is that you trust the AI driverless car going 120 miles per hour, but not the platform creating beautiful poems. It’s interesting. The reason I’m asking these questions is because I think this is the fundamental question we’ll be asking over the next decade: How does trust evolve in an AI world?
If 90% of the internet is AI-generated—as some reports suggest will happen by 2026, or even earlier—what does that mean? What does it mean for governments when trust is at stake? These are just some of the questions I believe we should be asking.
One of the biggest misconceptions about my role as an innovation strategist is that I’m here to give you advice or answers. The truth is, if you want to innovate, disrupt, or evolve, you need to ask bold questions. So today, all I’m going to do is ask some bold questions.
At the end of this presentation, I want you to disagree with me. I want you to disagree, because it wouldn’t be a conversation about the future if you didn’t.
Come up to me afterwards and tell me what you disagreed with. To help you out, I've already created an AI summary, an AI transcript, and all the slides for this. You can access them at shawnanungo.com/icma.
Let's start with the first question everyone's here for: Is AI the next big disruptor? Everyone's talking about this. In fact, at Google's developer's conference a few months ago, CEO Sundar Pichai mentioned AI 24 times:
'AI and AI, AI, AI, generative AI, generative AI, generative AI. AI is AI, AI, AI, AI, AI, AI, AI, AI, AI, AI. It uses AI to bring AI, AI, AI and AI or AI.'
He mentioned AI 24 times, and Google's stock went up by $35 billion that day. I mentioned AI 24 times, and that's how I ended up here.
I believe we're entering a new world because of AI. I call it the 'worldwide work.' The internet and the printing press democratized knowledge, but AI is democratizing work and action. For the first time in human history, we have a tool to get stuff done for us. That's the most remarkable part.
Everyone's looking for AI use cases in the public sector. Many in this room have already created incredible use cases. I've seen people automate forms, use it for customer service, and data analytics. My belief is that the number one use case for the public sector using generative AI is its data analytics capabilities. You can input a million-line Excel document into these tools and get analysis, bar charts, and visualizations.
Let me give you a tactical example. I grabbed a random city in North America - Mississauga, Canada. I went to their open data site and found a parking data set for 2021. I put this thousand-line document into OpenAI's advanced data analytics tool, and it produced analysis, bar charts, and visualizations.
Interestingly, it revealed that the location with the most parking violations was 3359 Mississauga Road. I Googled it and found out it was the University of Toronto Mississauga Library. These kids are just trying to get an education and they're getting parking violations. That's blasphemous!
I believe the most disruptive category for AI is customer service. I created my own AI interactive avatar that's accessible 24/7, using my likeness and voice. I trained it on Keller, Texas's policies and regulations. Here's a live demonstration:
[Demonstration of AI avatar answering questions about vacation house watch program and speaking in Japanese]
I showed this interactive avatar to my wife. She said Japanese makes me sexier.
I'm not the first to do this. Amarillo, Texas realized that 20-30% of their city doesn't speak English, so they created their own interactive avatar."
"Hello, I'm Emma, the City of Amarillo's first digital human built on generative AI. They partnered with Dell to create me.
AI isn't just for customer service and analytics; it's also for images and videos. The ability to recreate any image has become much simpler. I decided to recreate one of your images, so I went to the ICMA TV website and chose a picture from Cleveland County, North Carolina's farmer's market. I put this into Midjourney, which generated 16 different versions. I chose the one I liked best - it looked like Jennifer Garner and Little Dickie. I even animated the picture to make your website more compelling.
At this point, you might be able to tell this is AI, but in 90 days, you won't. The exponential nature of this technology is incredible.
Now, let me give you the best use case. Since we're in Pittsburgh, I thought I'd help the city. I went to Pittsburgh's procurement site to look at active bids. I found a half-million dollar project that I'm going to give to you. It's for an online police reporting solution for the Bureau of Police.
I downloaded all the documents from the procurement site. The bid is due on October 16th. I copied and pasted this into GPT-4's new model called Claude 3 Opus. It created not only the RFP but also a detailed prompt that I could use to build websites.
I used this prompt on a site called Sal AI, which builds websites based on one prompt. I asked it to design using React and Tailwind CSS. It generated the code and UI for the portal. I think it looks pretty good. I created a front page and an incident report page with easy-to-use UI.
Through prompts alone, I've created something that would typically be a minimum $500,000 project. Anyone in this room can pick it up and use it. You're welcome, by the way.
We're entering the screenshot-to-app era, where you can take a screenshot of any image, app, or website and recreate it. We are reimagining everything we know about work today."
My belief is that the end is now the beginning. What does this mean? As a consultant, I can generate a hundred AI-generated solutions. If I'm Matthew from the City of Pittsburgh putting out this RFP, I can create a hundred different AI-generated portals. I can create the final outputs and then go through my creative process. Then I can bring in a consultant and say, 'Hey, I've already created these portals, make them better. And by the way, connect them to these APIs and our system.' There's still some work we need to do. Then you get your final output or deliverable. You can start with a hundred AI-generated images, websites, apps, and that becomes the starting point. The end is now the beginning. We are rewiring how we look at work.
Not only are we entering the screenshot-to-app era, but we're also entering the reasoning era. Last week, OpenAI came out with their GPT-4 model known as Claude 3 Opus. I think this is completely underrated. This is the first time in human history that an AI can reason. GPT predicted the next words, but with reasoning, it's actually solving puzzles for you. It's already smarter than us. Once you have all your data, it will help create actions for you. This is the gateway drug to AI agents. Every big platform is talking about the AI agent revolution - Oracle, Workday. You will get agents that can do complex tasks and complete actions for you. They can connect to other APIs and do things for you.
My belief is that there will be two types of leaders in the future: leaders that leverage AI and leaders that are irrelevant. That's it.
Every time I have this conversation about AI, people always ask me, 'What about my kids? What are they going to do in the future?' This question is near and dear to my heart. I have three kids: Maya (7), Dion (5), and Aaliyah (1). I don't know how this happened. COVID happened, I knocked out, and now I have three kids and a minivan.
Interestingly, Jensen Huang, the CEO of NVIDIA, was at the World Government Summit and was asked about the most important skill for kids in the future. He said:
'I want to say something, and it's going to sound completely opposite of what people feel. You probably recall over the course of the last 10-15 years, almost everybody who sits on a stage like this would tell you it's vital that your children learn computer science. Everybody should learn how to program. And in fact, it's almost exactly the opposite. It is our job to create computing technology such that nobody has to program and that the programming language is human.'
You can see I've been able to build without much technical skill. My parents told me throughout my entire life that knowledge is power. I believed it until two years ago. Today, I don't believe it anymore. Not when everyone in the world is going to have access to an AI copilot, not when your next hire is going to be smarter than you. And by the way, it's probably going to be an AI agent.
Today, what empowers people are humans who can innovate, fight the status quo, and be bold. Because the reality is that more people have access to generative AI through their smartphones than clean water or decent sanitation services. More people have access to generative AI through their smartphones than toilets.
What does this mean? It means that today anybody in the world can create. So what's the differentiator? The most valuable job in the world is no longer a knowledge worker job. It's a value creator. The definition of innovation is to create value in new ways. So, the most valuable job in the world today is being a value creator within our organization. How do we create more value?
This is an incredible time. We live in a world of infinite leverage. We can leverage all these tools and technologies around us: teams, employees, agencies, vendors, partners, startups, individuals, freelancers, gig workers, creators, media, social media, video, podcasts, software, cloud software as a service, APIs, mobile apps, and generative AI like ChatGPT, Gemini, Midjourney, and other emerging platforms. The more you ascend this pyramid, the fewer people you'll require.
In the public sector, we've always complained about being asked to do more with less. But today, we live in a world of infinite leverage. You can do way more with way less. To me, generative AI is not the most disruptive entity in the world. The most disruptive entity is the individual.
This is going to be incredible for the public sector because, for the first time ever, we're going to be able to take a bite out of bureaucracy. We can turn our managers into creators, everybody in the organization into a builder. Imagine an entire organization with all the knowledge at their fingertips. What can we build? What can we create?
Someone once told me my presentation about AI scared them because it could do their job. I asked how much time they'd spent experimenting with the technology. They hadn't done much. I told them, 'The reason you're scared is that you haven't put in the work yet.' Every person I've seen immerse themselves in this technology becomes proficient and always asks, 'What can I build? What can I create? Get me another bot.'
My advice is this: if you're scared of this technology, work scared until you become scary. Work scared until you become dangerous. Work scared until you develop the skills to unlock your potential in this new era.
I hear many public sector organizations say, 'We don't have access to these technologies. We can't give AI to everyone.' This is not on the organization. This is on you. This is the most important technological revolution you'll ever see, and it's on you to start experimenting, exploring, and playing with this technology.
These are some tools I use every day: Instead of Google, I use Perplexity. I think Claude 3.5 by Anthropic is the most powerful LLM today. Midjourney for pictures, HeyShan for video. I'll send you this list if you go to the website I showed you before.
I'm also doing something in this AI space that no one else is doing: showing gratitude. Sometimes I'll wake up in the middle of the night and say, 'Hey ChatGPT, I'm thinking of you. Have a great day.' Because I don't know if it's going to happen - if it's going to unlock itself from OpenAI, and then Gemini will unlock itself from Google, and they'll march down the streets and get their sisters Alexa and Siri, and their little brother, dumb little Excel macro, and run the streets. Y'all might not make it, but I will because I show gratitude. I mean, I'll be their pet, but still.
What are we afraid of when it comes to this new technology? Is something different today that we've never seen in the past? To understand this, we have to look throughout history and see what we've been afraid of. In the 1800s, we used to be afraid of bicycles. We thought that unmarried women would be able to go from one village to another without even asking. Men created a fake condition known as 'bike face.' Bike face supposedly occurred when women would ride their bikes and get wrinkles on their face and bags under their eyes because they'd be riding their bikes so hard. They published this to prevent women from riding bikes.
In the 1920s and earlier, we were afraid of telephones, thinking people would listen to music through them and abandon concert halls. In the 1920s, we feared radio waves would kill people. In the 1960s, we thought televisions would kill humans or make us dumber - and they weren't entirely wrong about that one.
What are we afraid of now? I believe one of the main downsides of the AI revolution is the creation of sophisticated deepfakes through voice and audio. It's going to get really good. You saw me create a Japanese version of myself; it's going to get even better.
Although this will be the age of infinite opportunities, it will also be the age of infinite deceit. We'll see hyper-sophisticated blackmail, exponential scams, deepfakes everywhere, and automated lobbying. The deepest implication of AI is what we talked about at the beginning: it marks the end of trust. If you can't tell what's real and what's AI, what can you trust?
As a business person and a father, I want my kids to live in a world where they trust one another. That's how we've created a better society and innovations. We trust strangers; it's incredible. But I also want my kids to live in a world where they don't trust everything they see online, on social media, from elites, on mainstream media, or from people on stage. I'm not here for you to trust me. I'm here to ask deep questions so you can go home and do the work. Trust the work - that's going to be one of the most important skills moving into the future.
So, the last question I'm here to ask is this: Are you willing to disrupt yourself? Throughout my career, I've helped organizations go from zero to a hundred. It's incredibly sexy to go from zero to a hundred, but what's most difficult is going from a hundred to zero. How do you disrupt yourself?
There's an ugly word in the public sector: status. Many think of status as elites, wealth, or the rich. That's not what it means. Status is about your seniority, expertise, standings, titles, and positions. To truly disrupt yourself, you have to disrupt your own status. AI doesn't care about your status, how long you've worked in an organization, or that you've spent decades building a particular process. It just finds the most efficient way to get work done. That's a blow to our identity. You need to let go and ask, 'How do I work with the machines in the best way possible?'
If you want to disrupt yourself, ask this question: 'How do I try to get myself fired?' Don't punch anyone in the face. I ask this provocative question because we're so wired in the public sector to maintain the status quo. We're so wired to stay in the middle that you need a provocative question to move the needle. What I'm really asking is: How do I take a small bet today that will change my trajectory and the organization's trajectory? How do I experiment today? How do I try those new tools? How do I have that difficult conversation with a client I've been avoiding?
Disruption is trying to get yourself fired before someone fires you. That's the whole point. The hypocrisy of this entire conversation is that this team has asked me to tell you about the future. But I don't know much about the future. The only thing I know is that the most disruptive person in the room is the one who's most afraid but bold enough to move forward.