AI Everywhere: Transforming Our World, Empowering Humanity
- Mira MuratiSpeaker
- Jeffrey BlackburnHost
Key ideas
Murati explains how OpenAI learned to turn research into products and argues that safety work, staged releases and frontier regulation should develop alongside capability.
From a machine transcript of the audio. Quotes link to the moment they are said.
- 1
Capabilities are a hard place to start
Murati says OpenAI assumed commercial partners would turn GPT-3 into products through its API, then found that very hard for them, so it began building products itself. Her explanation is that you begin from a general capability rather than from a problem in the world.
- 2
Build safety alongside capability
Murati argues that safety, security and societal impact cannot be an afterthought and need to be built in alongside the technology. She says a smarter system is easier to direct, and that OpenAI is trying to predict a model's capabilities before training ends so that guardrails can be prepared.
- 3
Widen access in stages
For voice, Murati describes starting with experts and red teamers, building mitigations, then moving to about 1,000 users and on to larger groups as confidence grows. She says OpenAI does not let people make their own voices with the technology because it is still studying the risks and does not yet feel able to handle misuse.
- 4
Regulate the frontier, not the whole ecosystem
Murati says OpenAI has advocated more regulation for frontier models, where the risks are higher, while leaving room for smaller models from people with fewer resources. She favors regulation built on work already done in practice, plus more forecasting for systems still to come.
- 5
Nobody has measured the jobs impact yet
Murati says she is not an economist and that she does not think anyone really knows how many jobs will be created, changed or lost, because this is not being rigorously studied. She expects the tools to create a lot of value and names the open question of how it gets distributed.
In Mira’s words
“You're starting from capabilities. You're starting from a technology. You're not starting from what is the problem in the world that I'm trying to address.”
“For capabilities and safety, they're actually not separate domains. They go hand in hand.”
“If the nature of jobs really changes, then how are we distributing sort of the economic value into society? Is it through public benefits? Is it through UBI? Is it through some other new system?”
About this entry
- Published
- Jun 19, 2024
- Channel
- Dartmouth Engineering
- Type
- Video
- Length
- 51:42
- Language
- English
- Added
- Nov 2, 2026
- Source last checked
- Nov 2, 2026
What’s available here
Original source
Linked. Opens on YouTube.
Player on this page
Available. YouTube’s own player, loaded only when you press play.
Full text or transcript
Not reviewed. No license to show a text version here.
Our summary
Draft. Written by us with AI from a machine transcript of the video’s audio (Whisper) and checked against it. Machine transcripts can mishear; each quote links to the moment so you can hear it. Not yet rights-reviewed.
Ask questions with AI
Not reviewed. Off until there is licensed text to work from.