What Are We Thinking — in the Age of AI?

In November, 2022, OpenAI released ChatGPT 3 — a Large Language Model (LLM) — that dazzled and surprised even the people who produced it. The LLM’s impressive output quickly led to magical claims that Artificial Intelligence (AI) would write all of our software, cure cancer, solve climate change, and get rid of mold and mildew in bathrooms everywhere. Meanwhile, other claims (some of which new either) suggested that AI posed an existential threat to all humanity, or to testing and development jobs.

Only a few people — mostly in the research community, and hardly any in commercial software development — actually tested these claims skeptically and seriously. By and large, reports of problems and limitations were dismissed, and people reporting them were marginalized. Still, the problems were real, and couldn’t be ignored. When GPT-4o was released in May 2024 (shortly before this description was written), people experimented with it and published their experiences, whereupon the trickle of problem reports turned into a steady stream.

As of this writing, it’s not at all clear what the world will look like — economically and technologically — in the week of October 14, 2024. The drunken optimism may have been replaced by pragmatic sobriety, or by a severe hangover — or maybe climate change and cancer will be cured. We can’t predict the future. We can be reasonably sure of one thing, though: those who examine history, take scientific approaches, practice critical thinking, explore risk, and who test will have certain perspectives and advantages that aren’t available to others.

People who are skilled and dedicated to testing are a kind of defense early warning system for the business. In this keynote, Michael will challenge us — and himself — to critically assess our relationships with technology and hype — and our own beliefs.

Michael Bolton

Michael Bolton is a consultant and trainer who specializes in Rapid Software Testing (see http://www.developsense.com), a skill set and a mindset focused on high-value testing that is fast, inexpensive, credible, and accountable. He also writes extensively about testing and software development on his blog (see http://www.developsense.com/blog).

He focuses on helping people to solve testing problems that they don't know how to solve, and helping them learn how to do that for themselves. One way he does this is through teaching Rapid Software Testing—one of the world's most highly regarded classes on software testing. He works in any development context: Agile, Lean, DevOps, Scrum, or more traditional development models.