The Pulse

4,971 transcripts, every word counted. What Silicon Valley talks about, year by year, and who said it on this show first.

The interview, measured

Computed from every diarized word: how much the hosts talk, how hard the questions press.

Host talk share18% in 2026
0%20%40%2015201720192021202320262015 · 23%2016 · 27%2017 · 31%2018 · 33%2019 · 34%2020 · 31%2021 · 28%2022 · 23%2023 · 19%2024 · 19%2025 · 20%2026 · 18%

Share of all words spoken by the host. It peaked at 34% in 2019; the guest gets more of the mic every year since.

Questions per minute1/min
00.81.52015201720192021202320262015 · 1.22016 · 12017 · 0.92018 · 12019 · 12020 · 12021 · 12022 · 1.12023 · 1.22024 · 1.32025 · 1.22026 · 1

The pace of interrogation: pure math over every question mark in the transcripts.

Pushback4.6/10
05102015201720192021202320262015 · 1.62016 · 1.92017 · 2.12018 · 2.12019 · 1.92020 · 1.92021 · 2.22022 · 2.92023 · 3.72024 · 4.62025 · 4.62026 · 4.6

How often claims get contested instead of accepted: from 1.6/10 in 2015 to 4.6/10. The interview got harder.

Follow up6/10
05102015201720192021202320262015 · 4.32016 · 4.42017 · 4.72018 · 4.82019 · 4.72020 · 5.12021 · 5.42022 · 5.82023 · 6.22024 · 6.82025 · 6.62026 · 6

Drilling into vague answers instead of moving to the next question on the list.

Setup questions41% in 2026
0%40%80%2015201720192021202320262015 · 69%2016 · 65%2017 · 62%2018 · 63%2019 · 65%2020 · 61%2021 · 60%2022 · 50%2023 · 43%2024 · 36%2025 · 37%2026 · 41%

Setups, compliments and promotional prompts as a share of questions: 69% in 2015 → 41% now. The bigger the show got, the fewer of them, except at the top of the prominence scale: Star Treatment →

Episodes measured89 in 2026
01252502015201720192021202320262015 · 1282016 · 1562017 · 1482018 · 962019 · 952020 · 992021 · 912022 · 1652023 · 2412024 · 1462025 · 1472026 · 89

Episodes with full speaker diarization. These feed the charts around this one.

pushback = contests claims · follow up = drills into vague answers instead of moving on · setup questions = setups, compliments and promotional prompts as a share of questions. Rated by an LLM with a published rubric; talk share and questions are pure math.

The show itself

00m10040m20080m2015201620172018201920202021202220232024202520262015 · 128 episodes2016 · 156 episodes2017 · 148 episodes2018 · 96 episodes2019 · 95 episodes2020 · 99 episodes2021 · 91 episodes2022 · 144 episodes2023 · 149 episodes2024 · 146 episodes2025 · 148 episodes2026 · 111 episodes2015 · median 25m2016 · median 27m2017 · median 27m2018 · median 31m2019 · median 34m2020 · median 37m2021 · median 39m2022 · median 50m2023 · median 61m2024 · median 66m2025 · median 75m2026 · median 77mepisodesmedian length

Every processed episode by year published. Bars count them, the gold line tracks the median runtime. The show got longer as it got bigger.

The edit index: how produced the show sounds, year by year

The same guests, measured two ways. On the podcast feed's produced audio the "um"s and false starts were edited out in post production. For years the median guest shows ≈0.7 um/uh per 1,000 words, a rate no spontaneous human achieves. On raw YouTube recordings the median is ≈10.8. Of 132 guests we've heard on both, 114 show more filler raw (median +17.4/1k). This is why the speaking style measurements only use unedited recordings.

Produced podcast audiomedian um/uh per 1k words
020402015201720192021202320262015 · 31.62016 · 15.72017 · 3.12018 · 0.92019 · 0.62020 · 0.22021 · 0.22022 · 02023 · 02024 · 02025 · 1.22026 · 0.9

2017–2021 sits at ≈0. Credit the editing suite, not superhuman fluency. The early shows from the Skype era (2015–16) went out barely edited.

Raw YouTube recordingsmedian um/uh per 1k words
020402022202420262022 · 16.72023 · 11.42024 · 13.52025 · 8.42026 · 7.9

Same show, same years, unedited source: how people actually talk. Both charts share one scale.

The buzzwords

efficiency 2,395 mentions
2015peak 223.7/M words2026
first said 2015-01-19 by Kent Godfrey @11:50
crypto → 1,456 mentions
2015peak 386.6/M words2026
first said 2015-02-20 by Frank Meehan @5:11
LLM → 1,134 mentions
2015peak 183.8/M words2026
first said 2019-10-04 by Jason Boehmig @17:55
secondaries → 1,013 mentions
2015peak 118.5/M words2026
first said 2015-03-19 by Sharon Wienbar @17:11
PLG → 643 mentions
2015peak 121.6/M words2026
first said 2020-01-24 by Rick Nucci @17:22
unicorn → 643 mentions
2015peak 66.8/M words2026
first said 2015-02-05 by Niko Bonatsos @4:47
moat → 497 mentions
2015peak 70.7/M words2026
first said 2015-06-18 by Alan Chiu @7:47
AGI → 486 mentions
2015peak 91/M words2026
first said 2016-09-07 by Harry Stebbings @19:16
vibe coding → 311 mentions
2015peak 101.1/M words2026
first said 2025-04-11 by Aatish Nayak @57:28
NFT → 223 mentions
2015peak 88.6/M words2026
first said 2021-08-02 by Mike Lazerow @32:58
metaverse → 74 mentions
2015peak 17.9/M words2026
first said 2017-04-17 by Josh Felser @21:37
burn multiple 33 mentions
2015peak 14.5/M words2026
first said 2020-06-15 by David Sacks @22:27
ZIRP 31 mentions
2015peak 7.7/M words2026
first said 2022-11-14 by Harry Stebbings @15:13
wrapper 28 mentions
2015peak 5.2/M words2026
first said 2023-08-18 by Richard Socher @21:55
Made with StarZero

Turn any episode into a week of clips.

This entire site, over 1,200 episodes 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.