JSON
product on 11 shows · 12 statements across 11 episodes · said 30 times in 21 episodes since 2013
the MAD Podcast 11 
Latent Space 10 
the a16z Podcast 4 
Top Founders 2 
the Y Combinator Startup Podcast 1 
Starter Story 1 
the Startup Ideas Podcast 1 
Lenny's Podcast 
Sourcery 
Big Technology 
TBPN
Mentions by year, every show
tap a year for its mentions
the MAD Podcast 11Latent Space 10the a16z Podcast 4Top Founders 2the Startup Ideas Podcast 1Starter Story 1the Y Combinator Startup Podcast 1
2026 3 mentions in 3 episodes 1 per episode
2025 6 mentions in 3 episodes 2 per episode
2024 12 mentions in 7 episodes 2 per episode
- High Agency Pydantic over VC Backed Frameworks — with Jason Liu of Instructor
- Language Agents: From Reasoning to Acting — with Shunyu Yao of OpenAI, Harrison Chase of LangGraph
- How To Hire AI Engineers (ft. James Brady and Adam Wiggins of Elicit)
- A Brief History of the Open Source AI Hacker - with Ben Firshman of Replicate
- How To Create Ai Agents From Scratch (CrewAI, Zapier, Cursor)
- Building The Database That Can Do It All | Tobie Morgan Hitchcock, CEO of SurrealDB
- He Did $1m last year selling South American Business Data to firms like CapIQ
- every mention in 2024, scene by scene →
2023 1 mention in 1 episode
2021 1 mention in 1 episode
2019 1 mention in 1 episode
2015 5 mentions in 4 episodes 1 per episode
- 10 Commandments for BI in Big Data, Shant Hovsepian, Arcadia Data (Data Driven NYC / FirstMark)
- David Luan, Dextro // Real-World Video Understanding (FirstMark / Data Driven NYC)
- Emil Eifrem, Neo4j // Graphs Are Eating The World (Hosted by FirstMark Capital)
- Bob Muglia, Snowflake // Navigating The Cloud (Hosted by FirstMark Capital)
- every mention in 2015, scene by scene →
2013 1 mention in 1 episode
every mention on every show, scene by scene, with the transcript →
12 statements about JSON, every show
Unclosed JSON syntax is the primary cause of model tool hallucinations
“The challenge with tool calling more and more seems to be that the companies want certain tool calling, which is a very sensitive thing to train. And because you're dealing with all of the JSON outputs, if it doesn't like close the end of the request in a very…”
Penn: 80% of early Claude instruction failures were invalid JSON outputs
“And what I saw was something like 80% of what people meant in the early days for this failure was Claude would not write the right JSON.”
Penn: Claude scores 99.9% to 100% on JSON schema formatting evals
“When we have versions of Claude, we actually run that eval and just check. I think at this point it's always a hundred percent or like 99.9. And so it's no longer a pain point.”
Field: Jailbreaking AI models requires conversational questioning rather than fancy JSON
“It's not like you have to do some fancy JSON thing.
Like, that's what people like to popularize it online with, but I mean, you can just talk to them and ask them questions.”
Roy: Chunking Agent Context Into JSON Drops Token Use 70-80%
“Instead of using a giant CSV file as context in the agentic workflow, let's turn it into a JSON. Actually, let's chunk it into multiple different JSONs and only call, like, getting that technical, but like, that dramatically would reduce the actual tokens cons…”
Isenberg: Formatting prompt text in JSON improves AI text rendering accuracy
“If you include the text in a JSON format, it's more likely to give you the right text.”
Brown: OpenAI's o3 already achieves practical program synthesis
“Like when people say program synthesis, like we're already there, like O three is program synthesis, but the programs are like JSON and Python.”
Pokrass: Use XML for structuring LLM inputs and JSON for parsing outputs
“I do think XML is very helpful for structuring prompts, whereas for parsing outputs maybe the story is a bit different. Like sometimes it's really useful to get outputs in JSON, so you can plug them directly into your application. But I do think the models wor…”
Schluntz: JSON Escaping Overhead Degrades LLM Performance Across the Board
“Like if you're trying to output a code in JSON, there's a lot of extra escaping that needs to be done. And that actually hurts model performance across the board. Where versus like if you're in just a single XML tag, there's none of that sort of escaping that …”
Harrison Chase says TypeScript yields better LLM tool-calling performance than JSON
“I saw some paper that used TypeScript notation instead of JSON notation for tool calling and it got a lot better performance.”
Schemaless JSON fails as companies scale beyond 50 employees
“It works well, like, if you're, like, a ten-person company, or, like, even to a fifty-person company, when you're, like, scaling to, like, thousands of people, like, you actually need a proper negotiation in between.”