StarZero All shows Try the podcast clipper →

Douwe Kiela

Research Scientist Director, Google DeepMind · 1 appearance on the record.

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

scientistfounderexecutiveacademic@douwekiela ↗LinkedIn ↗cl.cam.ac.uk/~dk427 ↗Wikipedia ↗

Douwe Kiela is best known for pioneering Retrieval-Augmented Generation (RAG) and co-authoring the foundational 2020 paper while at Facebook AI Research. He later served as Head of Research at Hugging Face and co-founded Contextual AI to develop enterprise-grade contextual language models and grounded AI agents.

15statements → 6claims → 2claims resolved → 3.8/5average certainty → 2.93/5average debate potential → 4.3/5argument clarity · the sources → 1said about them ↓

0 supported 0 partly supported 2 contradicted 4 not checkable as stated how the 6 claims stand · each chip opens the sources

2 predictions · 4 assertions · 4 opinions · 5 insights · every statement was checked. The predictions and assertions are the 6 claims: statements the public record can support or contradict. 2 are resolved, and 4 name no date, number or outcome precise enough to check. 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 Douwe argues and how the claims held up. Everything they said, and everything said about them, is in the tabs below.

Their most notable contradicted claim

Assertion Contradicted
Kiela: FAISS was the first vector database
“In the initial paper, we used a vector database or a face. So the words vector database didn't exist at the time. But so face was the first vector database.”
Douwe Kiela Mar 6, 2025 ▶ 21:28 Top AI Researcher on GPT 4.5, DeepSeek and Agentic RAG | Douwe Kiela, CEO, Contextual AI

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

4.3 / 5 directness 4.5 · coherence 4.7 · precision 4 · compression 3.8

redirected or did not address 1 of 13 assessed questions (8%). Watch them ▸

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 · 45.7 um and uh per 1k words

Measured by listening to the audio itself: 7,192 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 Douwe Kiela said on the MAD Podcast that made the record, most notable first. Filter by type, assessment or year in the ledger →

Opinion
Kiela: Core language model development is almost solved and plateauing
“It's not even really about language models anymore. That has almost been solved, right? That's kind of why you see things plateauing off a little bit as well.”
Douwe Kiela Mar 6, 2025 ▶ 11:20 Top AI Researcher on GPT 4.5, DeepSeek and Agentic RAG | Douwe Kiela, CEO, Contextual AI
Opinion
Kiela: Long-context LLMs are inherently incredibly wasteful
“Long context models are inherently incredibly wasteful. You're paying for all this compute, and that's maybe why some of the companies that are trying to really sell long context model, long context window models, they will make more money from that, right?”
Douwe Kiela Mar 6, 2025 ▶ 29:23 Top AI Researcher on GPT 4.5, DeepSeek and Agentic RAG | Douwe Kiela, CEO, Contextual AI
Insight
Kiela: DeepSeek proved frontier AI models can rely on synthetic data
“We have kind of an existence proof now that it's actually not that hard to do this and so you don't need to invest all that much in, in data, and you can use synthetic data and get a pretty good model out of that”
Douwe Kiela Mar 6, 2025 ▶ 3:12 Top AI Researcher on GPT 4.5, DeepSeek and Agentic RAG | Douwe Kiela, CEO, Contextual AI
Assertion Not checkable as stated
Kiela: DeepSeek's total development cost was at least 100x its $6M training
“So I would guess that they spent at least a hundred X The amount of that, that single training run, right?”
Douwe Kiela Mar 6, 2025 ▶ 9:58 Top AI Researcher on GPT 4.5, DeepSeek and Agentic RAG | Douwe Kiela, CEO, Contextual AI
Insight
Kiela: Fine-tuning cannot inject new knowledge into AI models
“One common misconception about fine tuning is a lot of people think that you can inject new knowledge into a model using fine tuning. And that is not true.”
Douwe Kiela Mar 6, 2025 ▶ 28:19 Top AI Researcher on GPT 4.5, DeepSeek and Agentic RAG | Douwe Kiela, CEO, Contextual AI
Insight
Kiela: Advanced RAG systems break down when scaling to a million PDFs
“You can build a very awesome demo on a couple of PDFs and things will probably work. But then you have to scale it up to a million PDFs, and then everything breaks down. And the reason for that is that a lot of these kind of advanced RAG systems still actually…”
Douwe Kiela Mar 6, 2025 ▶ 35:14 Top AI Researcher on GPT 4.5, DeepSeek and Agentic RAG | Douwe Kiela, CEO, Contextual AI
Prediction Not checkable as stated
Kiela: AI systems will probably never reach 100 percent accuracy
“When are we getting to a hundred percent accuracy? And I had to give them the bad news that probably never.”
Douwe Kiela Mar 6, 2025 ▶ 48:21 Top AI Researcher on GPT 4.5, DeepSeek and Agentic RAG | Douwe Kiela, CEO, Contextual AI
Opinion
Kiela: GPT-4o is already effectively a reasoning model via chain of thought
“I mean, you could argue that GPT-IV-O is also already a reasoning model. It just hasn't been trained on reasoning specifically, but, ah, it can do chain of thought, right? So if it can do chain of thought, it's basically already a reasoning model. It just hasn…”
Douwe Kiela Mar 6, 2025 ▶ 7:00 Top AI Researcher on GPT 4.5, DeepSeek and Agentic RAG | Douwe Kiela, CEO, Contextual AI
Prediction Not checkable as stated
Kiela: AI is heading toward specialized language models over generalists
“Where we're headed is that we will have more specialized language models.”
Douwe Kiela Mar 6, 2025 ▶ 13:31 Top AI Researcher on GPT 4.5, DeepSeek and Agentic RAG | Douwe Kiela, CEO, Contextual AI
Opinion
Kiela: Attention mechanism, not Transformers, was the real AI breakthrough
“So I would say, and maybe I'm biased because one of my best friends is, is on the original attention paper, but that was the real breakthrough. It's just like figuring out that you have this attention mechanism that actually allows you to yeah, to do a much be…”
Douwe Kiela Mar 6, 2025 ▶ 20:19 Top AI Researcher on GPT 4.5, DeepSeek and Agentic RAG | Douwe Kiela, CEO, Contextual AI
Assertion Contradicted
Kiela: FAISS was the first vector database
“In the initial paper, we used a vector database or a face. So the words vector database didn't exist at the time. But so face was the first vector database.”
Douwe Kiela Mar 6, 2025 ▶ 21:28 Top AI Researcher on GPT 4.5, DeepSeek and Agentic RAG | Douwe Kiela, CEO, Contextual AI
Insight
Kiela: Retrieval is the only way AI agents can handle proprietary data
“Really focused on retrieval because that's really the only way you get these agents to work on your data and your problems.”
Douwe Kiela Mar 6, 2025 ▶ 50:02 Top AI Researcher on GPT 4.5, DeepSeek and Agentic RAG | Douwe Kiela, CEO, Contextual AI
Assertion Contradicted
Kiela: FAIR was the first team to build a generative RAG model
“Why RAG became the way you name these things is because it's generative, right? So we were the first ones to have a generative model there.”
Douwe Kiela Mar 6, 2025 ▶ 18:38 Top AI Researcher on GPT 4.5, DeepSeek and Agentic RAG | Douwe Kiela, CEO, Contextual AI
Insight
Kiela: Enterprise RAG fails if complex document data is not properly extracted
“If you want to have a enterprise grade rag system, you are only as good as the data that goes into that rag system. So if you can't extract the data in the right way, so if you have like a sort of table structure and it has like nested information, you can't g…”
Douwe Kiela Mar 6, 2025 ▶ 31:43 Top AI Researcher on GPT 4.5, DeepSeek and Agentic RAG | Douwe Kiela, CEO, Contextual AI
Assertion Not checkable as stated
Kiela: Aligning language models needs only 100 examples via Anchored Preference Optimization
“So you can train on this when you only have like a hundred examples, you can really make a meaningful, meaningful difference.”
Douwe Kiela Mar 6, 2025 ▶ 40:20 Top AI Researcher on GPT 4.5, DeepSeek and Agentic RAG | Douwe Kiela, CEO, Contextual AI

The other half of the tape: Douwe Kiela's own voice is left out of every number here. Other people bring the name up 1 time in 1 episode on the MAD Podcast. every mention, with the transcript →

Who brings them up most Matt Turck 1

Every mention by year

tap a year for its mentions
0011112025episodesmentions
0112025episodes it came up in
000.50.5112025episodesmentions per episode
2025 1 mention in 1 episode

Appearances (1)

EpisodeDateSpeaking time
Top AI Researcher on GPT 4.5, DeepSeek and Agentic RAG | Douwe Kiela, CEO, Contextual AI Mar 6, 2025 35m
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