3min chapter

"The Cognitive Revolution" | AI Builders, Researchers, and Live Player Analysis cover image

Ignore Previous Instructions and Listen To This Interview with Sander Schulhoff, CEO of Learnprompting.org

"The Cognitive Revolution" | AI Builders, Researchers, and Live Player Analysis

CHAPTER

Generating Model Examples and the Impact of Self-Generated Examples

This chapter explores the concept of generating self examples to enhance model performance. They discuss generating few-shot examples and using Auto Chain of Thought for generating rationales. The chapter concludes with sponsor ads from NetSuite and Amniki.

00:00
Speaker 1
It's a punch line. Anyway, yes, poor Bruce had to go through two and a half hours and we never got to a punch line, but he coped. He was a good sport. But but anyway, so afterwards, we're talking in, you know, the back story of I quad, you know, I have this coin, I'm just listening to this for listeners. There's this really, really layered relationship between mathematical operators as knowers and this structure of mathematical operations and how they apply to physics and the nature of the physical world and how you can collapse those particular set and see sort of a structure of coherence that actually contains our mathematical physical insights in a new way. And so Bruce then sends me this super cool yin yang picture that I hadn't seen. Okay. And on one side is ontology. That's like in the black and then the white circle there is like epistemic. And then the other side is epistemology and then ontick. Okay. And so so the story here for folks that are, you know, in terms of following, so I get trained as a psychological scientist. All right. And Roy Baskar is completely correct. Psychological science in particular commits the epistemological epistemic fallacy. Okay. And what does that mean? Basically psychological science is well, how do you know, do you know based on folks psychology and intuition? Or do you know based on empirical epistemology, forage in the modern scientific revolution? This is the good way to know and this is the bad way to know and good knowledge is scientific. That's why there's an entire thing called clinical science in psychology. And basically all of your knowledge should be grounded in empirical science. That's the whole justification. Okay. So as if now and Roy Baskar, if you know where Baskar at all, as if the way in which you know now assures that what you know is valid, regardless of what you say to be real, i.e. ontology, like what do you mean by mind and behavior? If you can't answer your ontological references, okay, then you can't claim coherent scientific knowledge. All right. So, and in fact, I went back to my earlier stuff. I have a 2003 paper where I opened the Tree of Knowledge and theoretical Unification of Psychology. And I, you can see this in history, I start with, well, there's a problem of epistemology. Okay. And now I look back at, no, it is a problem of ontology epistemology and metaphysics people. There's a problem, the whole philosophy is all screwed up. But no, as a scientist, I thought about really philosophy was basically epistemology. So then I split and learned to split ontology off from epistemology and see, obviously, there's claims about what is real. And then there's how we know about them. Okay, that's split one. Split two, then, is I'm in the conversation with Alexander Bard, like late 2018. Okay. And I'm using ontology, blah, blah, blah. But my language is not clear, because I'm not clear whether I'm talking about beliefs about what is real, okay, and how we know about what's real versus what's actually real. And Bard's like, Henrikas, do you know the difference between the ontic reality and ontology? Okay. And I was like, yeah, kind of, but
Speaker 2
not really.
Speaker 1
All right.

Get the Snipd
podcast app

Unlock the knowledge in podcasts with the podcast player of the future.
App store bannerPlay store banner

AI-powered
podcast player

Listen to all your favourite podcasts with AI-powered features

Discover
highlights

Listen to the best highlights from the podcasts you love and dive into the full episode

Save any
moment

Hear something you like? Tap your headphones to save it with AI-generated key takeaways

Share
& Export

Send highlights to Twitter, WhatsApp or export them to Notion, Readwise & more

AI-powered
podcast player

Listen to all your favourite podcasts with AI-powered features

Discover
highlights

Listen to the best highlights from the podcasts you love and dive into the full episode