HOW TO
SLOW DOWN AI
There is no realistic way for one person to stop AI. But together, we can apply a little friction.
If we genuinely want to know what personal data an AI company holds about us, we can exercise our data rights and ask.
Once ChatGPT has written the request, send it to OpenAI's privacy team:
WHY
DO THIS?
AI is already global, commercially valuable and increasingly embedded into everyday life.
Development is taking place across companies, countries, universities, open-source communities and individuals. There is no realistic off switch.
But that does not mean its development has to happen without scrutiny.
That requires resources.
One request will not change the direction of AI, and we cannot honestly claim that every request directly takes resources away from AI research.
What it does create is a legitimate operational obligation, while allowing us to understand what these companies actually know about us.
More scrutiny. More transparency. A little more friction.
HERE'S THE
AWKWARD PART.
We love AI.
We want instant answers. We want technology to solve problems faster. We want things that previously took hours to take minutes.
AI is extraordinary precisely because it can give us those things.
AI helped create this very page.
We gave ChatGPT the idea.
The Robot helped research the subject, structure the argument, challenge claims that went beyond the evidence, improve the wording and write the code you are looking at now.
It even corrected the argument when we tried to claim that every Subject Access Request directly takes resources away from AI development.
That cannot be demonstrated, so the Robot told us not to say it.
What we can reasonably say is that legitimate requests create work that organisations have to process.
ChatGPT does not have personal beliefs and is not endorsing this campaign.
But there is something wonderfully ironic about using AI to help us ask harder questions about AI.
THE AI DOC:
OR HOW I BECAME AN APOCALOPTIMIST
Daniel Roher approaches AI as an expectant father trying to understand what rapidly advancing artificial intelligence could mean for his child's life and humanity's future.
He speaks with executives and experts from the AI world and explores both the enormous potential of the technology and the serious concerns surrounding increasingly capable systems.
That tension is very close to what we are talking about here.
WE CAN BE EXCITED
AND STILL BE CAUTIOUS.
AI could help accelerate science, medicine, productivity and our ability to solve difficult problems.
At the same time, researchers and others have raised concerns ranging from misinformation and economic disruption to misuse, loss of control and much more serious long-term risks.
Nobody can currently tell us with certainty exactly where increasingly capable AI ultimately leads.
It means uncertainty deserves scrutiny.
We use AI. We are excited by AI. We simply do not see why that means we should stop asking questions.
1,000,000 REQUESTS
KNOW ABOUT US?
One person exercising their data rights will not stop AI.
But imagine a million people independently deciding that they genuinely want to understand what information AI companies hold about them.
What exactly do you know about us?
WE MAY NOT BE ABLE
TO STOP AI.
But we can understand it better.
We can exercise our data rights. We can ask what information is being held about us. We can examine the answers. And we can keep asking questions.
READ.
QUESTION.
If AI's most alarming predictions ultimately prove wrong, that will be an excellent outcome.
Until we know, perhaps there is value in making the journey a little slower.
This page was built with help from the Robot. 🤖
General information only and not legal advice. Data-protection rights and an organisation's obligations depend on applicable law and individual circumstances. Subject Access Requests should be genuine requests concerning your own personal data.