This is the second of a three-part series about key learnings from this year’s South by Southwest (SXSW) technology conference from our colleagues Becky Cole and Adrienne Schafer.
Artificial intelligence (AI) has been a part of our world for decades, but only in recent years has it come into its own. As with many technologies, the cultural moment and user experience had to align before AI could reach its full potential. (Remember video chat in the 90’s?) It’s becoming increasingly clear that AI will have a profound impact on nearly every aspect of our lives. From healthcare and finance to entertainment and transportation, the possibilities for this technology stretch the imagination. As evidenced by the many sessions at SXSW 2023, it’s clear that AI is no longer a novelty, but a critical tool that will shape the future.
At a SXSW keynote session with Greg Brockman, the president of Open AI, there was much speculation on how AI will shape the future. Brockman wasn’t out to oversell the potential or impact of ChatGPT. Every time the audience tried to steer down a dark alley with questions about morality, content rights, capitalist interests and lost jobs, he painted a rosy picture of a world where ChatGPT remains a technology that humans control and not vice versa.
Will AI Take Our Jobs?
So, is AI coming for our jobs? Brockman pointed out that we haven’t automated a whole job yet – not even with machines and physical tasks. Humans have consistently stayed involved, though the level of involvement may change. Brockman said, “I think the real story here, in my mind, is amplification of what humans can do.” Do we bemoan no longer having to wash dishes or clothes by hand? In an ideal future, AIs can do the work humans might not have wanted to do and add personalization. In a session with Disney Parks, Experiences & Products Chairman, Josh D’Amaro, talked about the evolution of AI and how they are now backwards engineering their process with robots. He simply stated “we build machines and learn from them” after showcasing their newest robot roller skating and performing tricks. They didn’t seek to build a robot that could elicit emotion but it was a surprising finding that shifted the project.
In a panel focused on AI and journalism, there was a mix of responses to whether journalism jobs would soon be done by computers. One idea was that jobs that aren’t currently being done, could be done by AI. Perhaps a small business could never have afforded to hire a graphic designer, or a small newsroom couldn’t pay someone to cover the local politics beat. An AI could in theory listen to a recording of a town hall meeting, take notes, and write a first draft, allowing a small group of people to review and publish local news that was perhaps not otherwise financially feasible.
Is the Accuracy of AI Improving?
Laurie Segall of Dot Dot Dot Media, led the conversation with Brockman, and she pointed out that right now, ChatGPT answers “like a drunk frat guy” confidently stating completely incorrect things. Brockman acknowledges that this is a problem and they’re working on it. Although he also says less perfect is a good thing. That means we have less trust in the tool and are conditioned to question it. Brockman says the team is thinking about ways to share information with people about how certain an AI is that an answer is correct. We’ve already seen accuracy improve quite significantly from version 3.5 to 4.0. (see the chart below)
What Morality Questions does AI Create?
There are probably enough questions on morality and AI to create a college course on it. Just a few questions include, who gets to moderate, what info should and shouldn’t be included, what about copyright? Brockman’s response was to talk about how Chat GPT is a core technology, the tools built on top of that technology will have to sort through ethics but it’s possible people can create their own personal AIs thus democratizing control of the tool and circumventing, large, centralized ownership controlling interest. Everyone gets to make their own rules, is an interesting response, especially considering how well social platforms have done with content moderation. If anything, they’ve proven it’s no small or easy task.
Across AI-focused sessions at SXSW, people were both excited at the possibilities and skeptical about the outcomes. How our lives are ultimately affected will depend a little on the technology and likely a lot on how we decide to structure society in response. Could universal basic incomes be implemented as an offset to AI taking jobs? Will AI sit quietly and just help humans work to the best of their abilities? Will government regulations prevent consolidated AI monopolies? We’re in for some big changes in the coming years.
The future of AI is full of possibilities, and it’s clear that we are at a critical juncture in shaping its impact on our society. As we move forward, we’ll need to consider not only the potential benefits but also the potential pitfalls and to be proactive in addressing them. There is a lot of work needed to ensure that AI serves the greater good and enhances human potential rather than detracting from it. At SXSW 2023 the sense was the future of AI is in our hands, and it’s up to us to steer it towards a positive and equitable future.
