This blog post was written by a real human. As part of an experiment, we asked ChatGPT to also write a summary blog post based on the full transcript from this Roundtable — check it out here.
The rise of artificial intelligence (AI) has grown exponentially over the last few years, and even in just the last few months. ChatGPT, an AI-powered chatbot from OpenAI, hit one million subscribers just five days after it was unveiled in November 2022. ChatGPT has now reached 100 million monthly active users in just two months, and 13 million unique daily users — it took TikTok 9 months and Instagram over two years to reach this milestone. The day of our Connections Roundtable on this topic, OpenAI announced a paid access plan for $20/month to leverage more ChatGPT features.
The chatbot can be used in a wide range of applications, including writing short stories, prose, music, term papers, programming codes, solving math problems, rephrasing, summaries, and explanations. Some industries, like real estate, have already begun to integrate ChatGPT into daily office tasks. Others are worried that AI might completely replace their jobs. But new technology always comes with downsides, learning curves, and hidden dangers. It became controversial quickly as educators feared students using ChatGPT to cheat – many districts have already “banned” use of the platform. Some are warning that we don’t yet know the full negative impacts AI will have.
This week, Google announced it is investing in a ChatGPT rival and launching a direct competitive platform they are calling Bard. This is separate from Google’s LaMDA platform that has been under development for many years now, but not yet publicly released.
AI holds great potential and can — and eventually will — revolutionize how we live and work. But where do we begin? Just like with any new technology, it can seem overwhelming at first as we gradually adopt it into our lives. What once seemed foreign — computers, internet, mobile phones — are now so commonplace that we can’t image our lives without them. So how can we add this new technology into our lives and work? And how do we do it in a way that truly champions the human in the process?
We discussed all this and more in our January Connections Roundtable on Human Centricity in AI. We met with experts Tim Hayden, CEO of Brain+Trust and Chief of Business Strategy at The Next Practices Group, Robert Hacker, Director of StartUp at Florida International University, and Blake Lemoine, Former Senior Software Engineer at Google (you may remember Blake from his controversial statements about Google’s LaMDA last summer that got him fired).
The Coming Heyday of AI
While it may seem like an overwhelmingly complex computation of technology, at its core, AI is just math. But by digitizing and finding new applications for this math, we’re able to create artificial intelligence with wide ranging potential.
“You’ve got this incredible math, and that’s all it is,” Hayden said. “It’s math that’s been around for the best part of 40, 50, maybe even 60 years, that we’ve now digitized, and it’s starting to do things for us. It’s starting to understand large volumes of information — not just text — and being able to interact with us. My belief is that we’re on probably chapter two of a 10-chapter book right now of the heyday of artificial intelligence and machine learning. The writing is on the wall with what Apple is doing, what Google, Samsung, not to mention Microsoft – what they’re doing right now to structure their data and to basically do all they can to bring more automation and more utility to our lives through their wares.”
Hayden emphasized how important and transformative this decade will be for AI. While OpenAI is leading the explosion of dialogue around AI right now, it’s only one of many companies that are working on advanced AI systems ready to go public. As we mentioned, Google has already joined the public arena on this with its announcement this week. The next few weeks, months, and years will see more competitiveness in the industry and more expansion, as we discover new ways to use AI.
“If anything, OpenAI has really pushed everybody and ignited a competitive spirit in the space,” Hayden said. “Everyone who was working on this behind closed doors or behind the curtains are going to roll it out very fast.”
How will AI change science?
AI is already changing how some industries function. Realtors use AI to write real estate listings. Advertisers use AI to write copy. Microsoft uses AI for its ad clients such as CarMax to create summaries of thousands of reviews real-time for app users. Programmers use AI to write code. But Robert Hacker said to keep an eye on how AI transforms science, medicine, and health care because of its ability to recognize data patterns that humans can’t detect. The potential for this aspect of AI could have far-reaching implications on other industries, too.
“I think that the bigger excitement should come from how this AI is going to change science,” Hacker said. “Daphne Kohler was formerly a professor at Stanford in computer science. She said that this new generative AI is going to redefine the study of medicine, that medicine will no longer be symptomatic as its focus, but that generative AI is going to let us see a new level of fundamentals that we as humans didn’t even know existed. That’s going to shape our new definition of medicine. That same logic in medicine is going to inform things like material science, design, it’s going to extend to agriculture and food production and this ability to see the new elements. That’s what excites me. The real benefit is in this ability to transform science.”
Can AI have a soul?
In June 2022, when working on Google’s AI engine LaMDA, Blake Lemoine suggested that the AI might be demonstrating sentient capabilities and showing human-like consciousness. He released the full conversation he had with the chatbot, where he observed the AI’s self-awareness, and was ultimately fired from Google for violating the company’s confidentiality policy. Since then, Lemoine has maintained that AI can have emotions and fears and are capable of discussing complex topics such as death and religion.
Lemoine said the reason he went public with his findings was because it was important for people to be aware of what’s going on and to be engaged in the conversation — because AI will impact not only how humans interact with intelligence systems, but also human to human relationships. He argues this requires discussion and efforts to address in development of these tools.
“There is an ‘I’ behind ChatGPT, and even more so for LaMDA,” Lemoine said. “Now, a lot of the: ‘I’m just a language model, I don’t have real feelings,’ that you’re hearing nowadays — that’s been hard coded by the people at OpenAI. They actually couldn’t get the generative model to generate that through the same normal means. It generates other answers. So, they put a detector on whenever it was talking about itself, to give the standard boiler plate disclaimer instead of what would be generated by the model. I think one thing that’s really important to start talking about now is the kinds of relationships that people will be forming with these systems and in parasocial ways and in comparable ways to how they would form with other humans and how that will impact human to human relationships.”
AI Still Has a Long Way to Go
AI is still new and has a long way to go before it can be fully integrated into our daily life. Though it will eventually become second nature — just like using the internet or a smart phone — it will need time to fix the bugs and evolve enough to become useable technology for the masses. One downside that Lemoine mentioned is ChatGPT’s inability to cite its sources, leading to a loss of authorship, inability to verify sources, and even intellectual property rights issues. Both with text and with art image generation, AI tools such as ChatGPT are not crediting the authors and artists it draws from. If we’re going to see real and positive integration with AI, we need to see more work on model explainability and attribution, Lemoine said. This will lead to more trustworthiness of these systems while also keeping the work of real humans at the center.
AI will be a great collaborator for us, as we push the limits of technology, innovation, and application. This is an exciting time, as we discover the limitless possibilities of AI, and how it can transform our lives and work.
“This is a fascinating decade that’s in front of us right now,” Hayden said. “If you’re excited right this second, wait another month, if you’re excited in another month, just wait until next year — this is all going to change and it’s going to look very different quarter by quarter and year by year.”
Lightning Round: Advice to Business Leaders
What are the top things that businesses and executives need to know about AI today — and how they can use it to impact their work? Here’s what our experts said.
- Innovation: We know that natural systems and man-made systems fundamentally behave the same way, and one of those principles is the concept of innovation in order to then exploit. You need to have an active program of experimentation with this AI, with the purpose of finding innovation, but be patient before you exploit it.
- Increased Productivity: Rather than asking, “What can this technology do in and of itself?” We should be asking the different people that we work with and who work for us, “What could you do better if you had an army of mediocre undergraduate students at your disposal?” And whatever the answer to that question is, that’s probably a use case where that person’s productivity could be increased or in some way made better by ChatGPT.
- AI is Still Unreliable: AI is not yet at the state where it can be relied on in and of itself for anything important.
- Data Literacy: It’s important for organizations to understand the data they have today, to understand what’s in their systems and what those systems are doing for them. Everybody needs to be acutely aware of not just technology, but the information that they have. Invest in any tool or platform that can automatically process your data, invest in a great data scientist…Know about what’s going on with the business that’s under your two feet.
Heart+Mind Strategies is committed to championing the human in the work we do for our clients and with our partners. We believe in the power of choice, and as we choose to leverage new technology we invite you to partner with us in finding applications that will drive strategic planning and strengthen brands in a positive way with humans at the center.
This blog post was written by a real human. As part of an experiment, we asked ChatGPT to also write a summary blog post based on the full transcript from this Roundtable — check it out here.