Artificial intelligence (AI) has fascinated humanity for decades, being explored and developed ever since Logic Theorist was made public in 1956. From that moment on, researchers and engineers have been trying models that are better at replicating human cognition, all of them with different grades of success. Today, the public and the media would make it seem that the perfect AI already exists, given their reactions to models like DeepMind and ChatGPT. Is this the case?
Back when machines were first introduced during the industrial revolution, there was widespread concern that they would lead to massive unemployment and social unrest. While there were certainly some negative consequences in the short term, machines ultimately led to an increase in the standard of living for people all over the world. In a similar way, every new form of automation ever since has continued to make people wonder if human workers were about to become obsolete.
The hype around the latest AI advancements is nothing new. Over the years, the technology has gone through several major milestones, each one leading to a renewed sense of excitement, usually accompanied by fear of its potential to replace human workers. From DeepBlue beating Garry Kasparov to Eugene Goostman becoming the “first” AI program to pass the Turing test, most of us have experienced several of these hype cycles in our lifetime.
It is true that today, AI is more powerful than ever before. Recent models like ChatGPT, DALL-E, Stable Diffusion, and Midjourney have not only gotten extremely good at mimicking human capabilities, but they have also become accessible to anyone with $10 to spare. While there are many reports of companies replacing workers with AI already, many others have opted to integrate them into their workflow while reducing their responsibilities and salaries accordingly.
OpenAI’s CEO Sam Altman specifically stated 3 months ago that “ChatGPT is incredibly limited” and that “it’s a mistake to be relying on it for anything important right now” just 3 months ago. Have ChatGPT limitations already been solved, or are companies and entrepreneurs rushing and taking a huge risk doing so? For most of us, it would be difficult to say.
AI’s current capabilities and applications were the topic of discussion in the “Can You Tell Whether this Headline Was Written by a Robot?” panel, which took place as part of this year’s edition of Grit Daily House at SXSW. Tech journalist Rob Pegoraro sat with serial SAAS entrepreneur John Sung Kim, CircleClick Media CEO Anne Ahola Ward, and HKUST Professor of Computer Science and Engineering De Kai to talk about what the latest AI advancements mean.
Their experiences with AI, what the right tasks for existing AI models are, how to figure out if specific content was AI-generated, and many other topics were discussed by the panelists during the 20-minute conversation. To learn more about what these experts think about the latest AI hype cycle, make sure not to miss the video on Grit Daily’s YouTube channel.