Person
1 episodes · 20 quoted claims
Andrew Ng said that almost two years before the talk he saw a debate over whether software counted as an agent.
“almost two years ago, I saw this rising trend in AI that a lot of people are excited about. But within the tech community, there's all this debate where some people write software and says an agent, others say, no, that's not an agent”
Andrew Ng said he ran an unpublicized campaign to get people to adopt the word 'agentic'.
“I actually kind of ran a campaign that I didn't publicize, but I did it anyway to try to get more people to just adopt the word agentic.”
Andrew Ng said marketers later applied the word 'agentic' to everything, which helped the movement take off.
“a few months later, a bunch of marketers would get a hold of this word and slap it as a sticker on everything in sight, and that helped the movement take off.”
Andrew Ng described agentic workflows as having an AI iterate through outlining, researching, drafting and critiquing.
“With agentic workflow, the idea is we can ask the AI to take a more iterative approach and say, first write an outline, then do some research and write the first draft, then critique it.”
Andrew Ng said agentic workflows work much better for tasks such as medical advice, legal advice, tariff compliance and writing code.
“from medical advice, legal advice, tariff compliance, writing code for a lot of different things, these agentic workflows work much better.”
Andrew Ng said a task that can be written as a standard operating procedure may be a good candidate for a multi-step agentic workflow.
“if you can write a standard operating procedure, like an SOP, that also may be a sign is we're seeing if you can codify the SOP in a multi-step agentic workflow.”
Andrew Ng said advice earlier in the year that people should not learn to code because AI will automate it will be seen as some of the worst career advice ever given.
“I think we look back on that as some of the worst career advice ever given.”
Andrew Ng said non-engineers in Silicon Valley teams who know how to code are starting to outperform those who do not.
“the marketers, HR professionals, the analysts, the finance professionals, the ones that know how to code, they're starting to run circles around the ones that don't.”
Andrew Ng said his children, aged four and six, occasionally use tablets with him present.
“my kids four and six, um, they do use uh uh tablets occasionally, but I'm there with them, right?”
Andrew Ng characterized the previous US administration's AI safety thinking as safety theater driven by lobbyists seeking regulatory capture.
“There was really safety theater driven by lobbyists, um, fear-mongering to try to create regulatory capture, anti-open source types of regulations, right?”
Andrew Ng said he worries American competitiveness will suffer if high-skill immigration is made harder.
“I really worry about American competitiveness if we make it harder for high skill immigration.”
Andrew Ng said he came to the United States as an undergraduate student at about seventeen years old.
“When I came to the United States as an undergrad student, I think I was like seventeen years old.”
Andrew Ng said China's recent ban on certain Nvidia chip imports signals movement toward independence from TSMC in Taiwan.
“it's been interesting to see China recently banned certain imports of Nvidia chips, which is a strong signal that China is moving toward independence from TSMC in Taiwan”
Andrew Ng described a data center as a machine that turns electricity into intelligence, or tokens, and said energy is a bottleneck.
“when you build a data center, that's a machine to turn electricity into intelligence, or turn electricity into, um, tokens.”
Andrew Ng said the mobile ecosystem has two gatekeepers, Android and iOS, and hoped AI will not end up with few gatekeepers.
“the mobile ecosystem is kind of uninteresting is because there are two gatekeepers, Android and iOS.”
Andrew Ng said China has pulled ahead of the US in releasing open-weight models, which anyone can download and use for free.
“China has really pulled ahead of the US. in releasing open weight models. These are models anyone in the world can download and use for free.”
Andrew Ng said he saw a statistic suggesting cumulative adoption of Chinese open-weight models may have surpassed that of US open-weight models.
“I think I saw a stat showing that cumulative adoption of Chinese open weight models, I think, is about to surpass or may already have surpassed cumulative adoption of US. open weight models.”
Andrew Ng said European regulators told him they wanted to lead in regulating AI, which he does not think creates competitive advantage.
“for a while, over the last few years, visiting some European regulators, I heard things like, we want to be leaders in regulating AI.”
Andrew Ng said people in Silicon Valley may underestimate how much distrust of AI exists across the United States.
“I think many of us may underestimate the distrust that a lot of people across the nation have for AI.”
Andrew Ng said he likes Claude Code for coding and is increasingly using OpenAI's Codex.
“I love, uh, for coding, I love cloud code and open increasingly using, um, open AI, uh, codex as well.”