In pursuit of our mission, we’re committed to making sure that access to, advantages from, and influence over AI and AGI are widespread. We imagine there are at the least three constructing blocks required with the intention to achieve these goals within the context of AI system behavior.^{[^scope]}

1. Improve default behavior. We wish as many users as possible to search out our AI systems useful to them “out of the box” and to feel that our technology understands and respects their values.

Towards that end, we’re investing in research and engineering to scale back each glaring and subtle biases in how ChatGPT responds to different inputs. In some cases ChatGPT currently refuses outputs that it shouldn’t, and in some cases, it doesn’t refuse when it should. We imagine that improvement in each respects is possible.

Moreover, we’ve room for improvement in other dimensions of system behavior equivalent to the system “making things up.” Feedback from users is invaluable for making these improvements.

2. Define your AI’s values, inside broad bounds. We imagine that AI needs to be a great tool for individual people, and thus customizable by each user as much as limits defined by society. Due to this fact, we’re developing an upgrade to ChatGPT to permit users to simply customize its behavior.

This may mean allowing system outputs that other people (ourselves included) may strongly disagree with. Striking the fitting balance here shall be difficult–taking customization to the intense would risk enabling malicious uses of our technology and sycophantic AIs that mindlessly amplify people’s existing beliefs.

There’ll subsequently all the time be some bounds on system behavior. The challenge is defining what those bounds are. If we attempt to make all of those determinations on our own, or if we attempt to develop a single, monolithic AI system, we shall be failing within the commitment we make in our Charter to “avoid undue concentration of power.”

3. Public input on defaults and hard bounds. One solution to avoid undue concentration of power is to provide individuals who use or are affected by systems like ChatGPT the flexibility to influence those systems’ rules.

We imagine that many choices about our defaults and hard bounds needs to be made collectively, and while practical implementation is a challenge, we aim to incorporate as many perspectives as possible. As a place to begin, we’ve sought external input on our technology in the shape of red teaming. We also recently began soliciting public input on AI in education (one particularly essential context during which our technology is being deployed).

We’re within the early stages of piloting efforts to solicit public input on topics like system behavior, disclosure mechanisms (equivalent to watermarking), and our deployment policies more broadly. We’re also exploring partnerships with external organizations to conduct third-party audits of our safety and policy efforts.