By playing, you agree to uListen's Terms and Privacy Policy and YouTube's Terms of Service.
Bloomberg OriginalsInside OpenAI, the Architect of ChatGPT, featuring Mira Murati | The Circuit with Emily Chang
Summaries, chapters, transcripts, and FAQs on this page are made by uListen, not YouTube.
At a glance
WHAT IT’S REALLY ABOUT
Inside OpenAI: scaling breakthroughs, safety risks, and the AI race
- Mira Murati describes OpenAI’s decision to release ChatGPT as a way to gather real-world feedback on risks, limitations, and how people use the system.
- Murati explains ChatGPT as a scaled neural network trained to predict the next word, while acknowledging major challenges around truthfulness, reliability, and hallucinations.
- The conversation highlights societal concerns including misinformation, the need to distinguish machine from human output, and uncertainty about how creators should be compensated as AI reshapes work.
- Reid Hoffman argues AI will be a major platform shift, predicts rapid integration into daily life, and says public discourse is too dominated by fear rather than hope.
- Both guests discuss governance and deployment: Murati supports trusted auditing authority, while Hoffman defends controlled API release and the nonprofit-plus-commercial structure as enabling safety and mission continuity.
IDEAS WORTH REMEMBERING
5 ideasOpenAI released ChatGPT largely to learn from public feedback, not just to ship a product.
She says the release decision came after the team had ChatGPT internally and reached a point where learning required real-world feedback about use, risks, and limitations, while also putting the technology into public consciousness.
ChatGPT’s core mechanism is next-word prediction, with capabilities emerging through scale and safety work.
She explains that the model is trained to predict the next word, and that scaling model size and data increases capabilities; she adds that increased investment in alignment and safety can make systems more reliable and safe over time.
Hallucinations are presented as a central, unresolved reliability risk.
Murati says a key worry is GPT-4’s ability to “make up things” convincingly, and stresses people need awareness that they cannot blindly rely on outputs.
The episode frames the future as human–AI collaboration with blurred output boundaries.
She says it’s important to distinguish machine output from human output, but also predicts collaboration will create “hybrid” outputs as people work with AI more.
AI’s labor impact is portrayed as both disruptive and generative of new job categories.
Murati says she doesn’t know how valuing everyone’s internet-created information would work in practice, but expects jobs to be lost and changed; she points to prompt engineering as an example of an unexpected new role.
WORDS WORTH SAVING
5 quotesSo they will convincingly make up things, and it requires, you know, being aware and just really knowing that you cannot fully, blindly rely on what, uh, the technology is providing as an output.
— Mira Murati
Sometimes when we don't know the answer to something, we will just make up an answer. We will rarely say, "I don't know."
— Mira Murati
I think we should be very careful in general with, um, putting very powerful systems, um, in front of more vulnerable populations.
— Mira Murati
Imagine a tutor on every smartphone for every child in the world. That's possible.
— Reid Hoffman
There's certainly a risk that, um, when we have these AI systems that are able to set their own goals, they decide that their goals are not aligned with ours and, uh, they do not benefit from having us around and could lead to human extinction.
— Mira Murati
SPEAKERS
Emily Chang
hostBloomberg journalist and host of “The Circuit with Emily Chang.”
Mira Murati
guestOpenAI executive (CTO) and key leader behind ChatGPT’s development and deployment.
Reid Hoffman
cameoVenture capitalist and LinkedIn co-founder; early OpenAI collaborator and Microsoft board member.
High quality AI-generated summary created from speaker-labeled transcript.
