StarZero All shows Try the podcast clipper →

Kanjun Qiu

Co-Founder and CEO, Imbue · 1 appearance on the record.

computed by AI from the episodes · how this works → · full disclaimer →

founderexecutiveinvestorauthor@kanjun ↗LinkedIn ↗kanjun.me ↗Wikipedia ↗

Kanjun Qiu is the co-founder and CEO of Imbue, an AI research lab developing reasoning and code-generating agents. She previously served as the first Chief of Staff at Dropbox, co-founded the startup Sourceress, and co-authored the computing education book Sew Electric.

12statements → 4claims → 3claims resolved → 100%fully supported → 4.17/5average certainty → 1.83/5average debate potential → 4.1/5argument clarity · the sources →

3 supported 0 partly supported 0 contradicted 1 not yet assessed how the 4 claims stand · each chip opens the sources

4 assertions · 1 opinion · 4 insights · 3 disclosures · every statement was checked. The predictions and assertions are the 4 claims: statements the public record can support or contradict. 3 are resolved, 1 is not yet assessed. Everything else (opinions, insights, what ifs, disclosures) can never be settled by the record, so it carries no assessment.

The record, in short

What the tape says about how Kanjun argues and how the claims held up. Everything they said, and everything said about them, is in the tabs below.

Their most notable supported claim

Assertion Supported
Kanjun Qiu: Language models trained without code perform poorly at reasoning
“Some companies have tried training language models with no code because they're like, oh, the product we're building, we don't need code. You know, it's a therapist or it's a question answering thing and we don't need code. So we should take code out of our tr…”
Kanjun Qiu Oct 4, 2023 ▶ 17:08 Building Human-Level AI Agents: Imbue CEO Kanjun Qiu on the Road to General Intelligence

Expressed certainty vs assessment result

none yet certainty 1
none yet certainty 2
none yet certainty 3
100% certainty 4
100% certainty 5

weighted support: a fully supported claim counts one, a partly supported claim counts half. Each filled bar is clickable and opens exactly those claims; "none yet" means nothing said at that certainty level has resolved yet

Argument clarity: do they answer the question? how? →

4.1 / 5 directness 4.1 · coherence 4.3 · precision 3.8 · compression 3.6

answered every one of 9 assessed questions directly

This is a score against a rubric. It is not a rank. Every host question → answer exchange is scored with names hidden on directness, coherence, precision and compression, 1–5 each, on meaning alone: disfluencies are ignored, and only raw unedited episodes count. This is the score that measures thought. Every scored exchange, scores shown → · The rubric and its checks →

How they sound: speaking style how? →

249 words/min while actually speaking · 42.4 um and uh per 1k words

Measured by listening to the audio itself: 6,322 words across 1 episode of raw-level tape, transcribed verbatim with every um and uh kept, each one attributed only where the alignment onto our timed stream is unambiguous. These are measurements of speaking style. We do not rank them: across this corpus, fluency and argument quality are nearly uncorrelated (ρ≈0.2), and smooth talking does not signal clear thinking. How it's measured →

Everything Kanjun Qiu said on the MAD Podcast that made the record, most notable first. Filter by type, assessment or year in the ledger →

Insight
Kanjun Qiu: AI emergent capabilities are artifacts of evaluation metric design
“If your metric is smooth you actually see slow performance improvement over time, and if your metric is relatively discrete or not smooth, that's where you see the emergence, and it's actually more about the evaluation metric Than about the emergence of the ca…”
Kanjun Qiu Oct 4, 2023 ▶ 21:11 Building Human-Level AI Agents: Imbue CEO Kanjun Qiu on the Road to General Intelligence
Opinion
Qiu: AI is currently in an un-reliable analog computer phase
“I think today, you know, we're kind of in the analog computer phase of AI, where You know, ah, these systems, they output things. Errors compound. They're not very reliable.”
Kanjun Qiu Oct 4, 2023 ▶ 14:02 Building Human-Level AI Agents: Imbue CEO Kanjun Qiu on the Road to General Intelligence
Assertion Supported
Kanjun Qiu: Language models trained without code perform poorly at reasoning
“Some companies have tried training language models with no code because they're like, oh, the product we're building, we don't need code. You know, it's a therapist or it's a question answering thing and we don't need code. So we should take code out of our tr…”
Kanjun Qiu Oct 4, 2023 ▶ 17:08 Building Human-Level AI Agents: Imbue CEO Kanjun Qiu on the Road to General Intelligence
Insight
Qiu: Psychological trauma can be conceptualized as computational overfitting
“Trauma is overfitting and actually a good way to overcome trauma. If we look at a lot of therapy It's like by giving more data to me and helping me access the overfit parts and then give them more data.”
Kanjun Qiu Oct 4, 2023 ▶ 40:08 Building Human-Level AI Agents: Imbue CEO Kanjun Qiu on the Road to General Intelligence
Insight
Kanjun Qiu: Self-supervised learning mirrors how humans naturally learn
“Self-supervised learning is interesting because it's how humans learn, you know, most of the time we go around We like Google stuff. We learn on our own. We're not giving supervised labels all the time.”
Kanjun Qiu Oct 4, 2023 ▶ 3:51 Building Human-Level AI Agents: Imbue CEO Kanjun Qiu on the Road to General Intelligence
Disclosure
Kanjun Qiu: 95% of Imbue's work is engineering, not research
“So the company is mostly engineering. You know, I think something people, when people think about AI, they're like, oh, it's research, but really, 95% of the work that we all do is engineering work.”
Kanjun Qiu Oct 4, 2023 ▶ 5:00 Building Human-Level AI Agents: Imbue CEO Kanjun Qiu on the Road to General Intelligence
Insight
Qiu: Current AI models are book smart but lack street smarts
“The way to think about today's models is they, they're like book smart, but not street smart. They're like a person who's read the whole internet, but they've never done anything in the world.”
Kanjun Qiu Oct 4, 2023 ▶ 8:48 Building Human-Level AI Agents: Imbue CEO Kanjun Qiu on the Road to General Intelligence
Assertion Supported
Kanjun Qiu: Imbue operates a cluster of 10,000 Nvidia H100 GPUs
“And then NVIDIA is also part of the round because we have a very large GPU cluster of 10,000 H 100, and it's very helpful to have NVIDIA as part of that.”
Kanjun Qiu Oct 4, 2023 ▶ 28:34 Building Human-Level AI Agents: Imbue CEO Kanjun Qiu on the Road to General Intelligence
Disclosure
Kanjun Qiu paid for MIT by designing high-frequency trading algorithms
“I paid for MIT by trading high-frequency trading algorithms, so by designing high-frequency trading algorithms”
Kanjun Qiu Oct 4, 2023 ▶ 1:40 Building Human-Level AI Agents: Imbue CEO Kanjun Qiu on the Road to General Intelligence
Disclosure
Qiu: Imbue pre-trains 100B+ parameter models from scratch
“We pre-train our own large model. So, you know, very large hundred billion parameter plus models from scratch with our own data.”
Kanjun Qiu Oct 4, 2023 ▶ 22:12 Building Human-Level AI Agents: Imbue CEO Kanjun Qiu on the Road to General Intelligence
Assertion Not publicly verifiable
Imbue used plasma physics principles to build its CARBS search algorithm
“Like the reason we have carbs is because Abe was a plasma physicist and he applied some ideas from plasma physics, physics to make this local search algorithm work well.”
Kanjun Qiu Oct 4, 2023 ▶ 32:39 Building Human-Level AI Agents: Imbue CEO Kanjun Qiu on the Road to General Intelligence
Assertion Supported
GPT-3 lead authors Tom Brown and Ben Mann were Qiu's housemates
“And out of that house our housemates, Tom Brown and Ben Mann are the first two authors on GBT-III.”
Kanjun Qiu Oct 4, 2023 ▶ 35:19 Building Human-Level AI Agents: Imbue CEO Kanjun Qiu on the Road to General Intelligence

Appearances (1)

EpisodeDateSpeaking time
Building Human-Level AI Agents: Imbue CEO Kanjun Qiu on the Road to General Intelligence Oct 4, 2023 30m
Made with StarZero

Turn any episode into a week of clips.

This entire site, over 400 conversations transcribed, diarized, checked and made playable, runs on the StarZero media pipeline. Drop in your own episode and the podcast clipper finds the moments worth sharing, cuts them, captions them, and reframes them for every feed.

Try the podcast clipper Learn more