People who want to slow AI down get called doomers. People who want to keep going get called reckless. These labels flatten what people actually believe and let us dismiss them without listening to their arguments.
Someone can see huge benefits in AI and still want to slow down. Someone can want to keep going while caring deeply about safety. They might share the same values but disagree about the risks and what to do. We should try to understand those differences before judging each other. Strong disagreement doesn’t require assuming that someone is indifferent to safety or progress.
Motives and incentives deserve scrutiny, too. But examining them shouldn’t replace engaging with an argument. Someone with a financial stake can still make a valid point, just as someone concerned about catastrophic risks can have reasons worth taking seriously. Fair criticism addresses what a person actually believes and explains why we disagree.
Pace or Pause exists to make that kind of engagement easier. It brings together influential voices on AI so readers can understand where they agree, where they disagree, and the reasoning behind their positions.
AI executives, researchers, politicians, advocates, and prominent influencers shape public opinion and decisions about AI. Their positions deserve closer examination than reputation alone allows. What benefits do they see? What risks concern them? What evidence are they relying on? And what do they actually think we should do?
A stance category gives readers a rough idea of where someone stands. It can’t explain all their reasoning or qualifications. It’s a starting point, not a complete picture of what they believe. Summaries should make the debate easier to navigate without making anyone sound more extreme or certain than they really are. When a position is unclear or mixed, the site should explain that ambiguity rather than force it into a category.
Original sources are essential to this approach. Even a fair summary leaves things out. Readers should be able to examine the context and nuance for themselves and decide whether my interpretation accurately represents what someone believes. Most people won’t read every source, but they should have a clear path to dig deeper.
Bringing positions together can also reveal common ground that gets lost when we encounter statements separately. People who disagree about the pace of development may still want many of the same outcomes. Seeing the risks they identify, the evidence they rely on, and the actions they recommend helps explain why they reach different conclusions.
When thoughtful people with relevant expertise disagree, readers don’t have to pick a side immediately. The disagreement may come from different evidence, assumptions, or values. Understanding its source can clarify what’s at stake, which questions remain open, and what evidence might change our minds.
I have my own position. I align most closely with Continue. I’m a techno-optimist: I believe AI has enormous potential benefits for society, and I worry that slowing development delays those benefits reaching people. I see alignment as a major challenge, but I’m optimistic that we can solve or manage it.
That optimism comes with conditions. I support safety testing, independent evaluations, and meaningful transparency that still allows companies to protect proprietary information. I believe companies should be held responsible for releasing dangerous models, alongside a limited set of safeguards. Strong evidence that models were causing serious harm or becoming harder to control, while safety testing failed to keep up, would make me reconsider. My optimism depends on our ability to manage the risks.
I’m sharing this so readers know where I’m coming from. The site’s purpose is to help them reach their own conclusions. My beliefs shouldn’t affect how honestly I present someone else’s position, and evidence that challenges my optimism deserves the attention its strength and significance warrant.
That commitment includes being willing to correct my work. If someone points out that I’ve misrepresented a position, I want to examine the evidence and correct any mistake openly. Getting someone’s position right matters more than defending my interpretation.
Readers don’t have to change their minds for this site to be worthwhile. I hope they leave understanding other people’s positions better, less quick to dismiss them, and with a clearer sense of the arguments and more thoughtful questions.
When we encounter a position we disagree with, we can ask: What evidence or experiences led this person here? What might I be missing? We can still disagree strongly, but our disagreements should be informed by the actual benefits, risks, and choices we face.
Explore the evidence and reach your own conclusions, while staying open to what you might be missing. You don’t have to agree with me or anyone else.