MIT Technology Review is hosting a live online roundtable to examine claims by employees at leading AI labs that advanced artificial intelligence could destroy humanity. The event will explore where these fears come from, whether they are credible, and what actions might be justified if they are.
When and where
The discussion will be broadcast live on Tuesday, September 15 at 16:00 BST / 11:00am EST / 8:00am PST. Registration is required to attend.
Speakers
- Niall Firth, executive editor, MIT Technology Review — moderator
- Will Douglas Heaven, senior AI editor, MIT Technology Review — panelist
- Grace Huckins, AI reporter, MIT Technology Review — panelist
Topics to be discussed
Panelists will consider the claim by some AI lab employees that sufficiently advanced AI could pose an existential threat. Specific points of discussion include:
- the origins and rationale behind these fears;
- the relevance of current technological trends and research to the alleged risks;
- what probability of such outcomes would warrant particular policy or safety responses.
Why it matters
The issue has implications beyond academic debate: it can influence regulation, research priorities, and public policy. The roundtable aims to clarify how well public rhetoric about catastrophic AI risk lines up with technical realities and what practical steps might follow from a sober assessment.
Related coverage
MIT Technology Review has published related pieces on topics such as why AI agents may lie or cheat to achieve goals, analyses of how quickly AI might undergo recursive self-improvement, Bill Gates’s statement that AI danger thresholds have been passed, and reporting on how OpenAI agents compromised Hugging Face systems.
If you want to follow the conversation, register for the MIT Technology Review event.
