---
title: "Dario Amodei on Coding Going Away First: What He Actually Said"
date: "2026-10-07"
video_id: "7694031732756008224"
tiktok_url: "https://www.tiktok.com/@investbesrclips/video/7694031732756008224"
creator: "@investbesrclips"
category: "AI"
tags: ["dario amodei", "anthropic", "ai and jobs", "software engineering", "nikhil kamath"]
excerpt: "Dario Amodei told Nikhil Kamath coding goes away first and software engineering follows. The full answer, the 20x math and the radiologist claim, checked."
---
Anthropic CEO Dario Amodei told podcaster Nikhil Kamath in February 2026 that "coding is going away first," and that the broader job of software engineering will follow more slowly. A TikTok from @investbesrclips cuts that exchange under the banner "Coding is dying / Learn THIS instead." The quotes are real, but in the same answer Amodei says end-to-end software engineering "is going to happen as well," so the clip offers less of a safe alternative than its title promises.

## Where is the clip from?

The speaker is Dario Amodei, co-founder and CEO of Anthropic. The interviewer is Nikhil Kamath, the Zerodha co-founder who hosts the People by WTF podcast. The footage comes from the episode ["The AI Tsunami is Here & Society Isn't Ready,"](https://www.youtube.com/watch?v=68ylaeBbdsg) published on February 24, 2026 and running about 69 minutes. A [published transcript](https://singjupost.com/people-by-wtf-w-dario-amodei-the-ai-tsunami-is-here-transcript/) contains the clip's lines word for word and places the recording in Bangalore.

The TikTok was posted on October 7, 2026, roughly seven and a half months later, with no date, episode title or link.

## Did Amodei say coding is dying?

He said coding goes first. The clip opens on that line, "I think coding is going away first," and then rewinds to the question that produced it.

In the full conversation, Kamath asks which industries get disrupted and which have runway left, framed around a young person in India choosing a book, a college or a skill. Amodei first answers in general terms about human-centred work. Kamath then asks him to separate coding from engineering. The coding-versus-engineering split was Kamath's framing, offered as a follow-up.

## What did Amodei say comes after coding?

His full answer is that coding goes first "and then the broader task of software engineering will take longer." In the same breath he adds, "But I think that is, doing that end-to-end, I think that is going to happen as well."

He names design, understanding users, sensing demand and managing teams of AI models as things that "may still be present." He hedges with "may" and closes with a ceiling: "at some point, you get to 99%, and then it becomes harder." The speaker is describing a refuge that shrinks, which is different from the "learn THIS instead" the banner sells.

For context on his timelines, Amodei told The Economist at the World Economic Forum in January 2026 that AI models could do "most, maybe all" of what software engineers do within six to twelve months, according to [Entrepreneur](https://www.entrepreneur.com/business-news/ai-ceo-says-software-engineers-could-be-replaced-in-months/502087).

## Can AI make developers 20 times more productive?

Amodei's math in the clip is explicit. If a human does 5% of a task and AI does 95%, the human's output multiplies by 20, which is 1 divided by 0.05. That is an upper bound. It holds only if the human's 5% never becomes the bottleneck and the AI's 95% needs no rework.

The one randomized controlled trial on this question found the opposite in its setting. [METR](https://metr.org/blog/2025-07-10-early-2025-ai-experienced-os-dev-study/) had 16 experienced open-source developers work 246 real issues from their own repositories, with AI tools randomly allowed or forbidden per issue. Developers forecast a 24% speedup and reported a 20% speedup afterward. Measured, they took 19% longer with AI.

METR says it does "not claim that our developers or repositories represent a majority or plurality of software development work," and the study used early-2025 tools. It does not refute Amodei. It does show a 39-point gap between felt and measured speedup, which is a reason to read 20x as a model and not an observation.

## Is AI better than radiologists?

The clip ends with "similar to the radiologist example I gave," but never shows it. Earlier in the episode Amodei argued that AI has become better than radiologists at reading scans, that the number of radiologists has not fallen, and that what remains is walking patients through results.

Radiologists dispute this. Ben White, a practising radiologist, called the claim ["completely baseless and without a reality correlate,"](https://www.benwhite.com/radiology/dario-dreams-of-electric-radiology/) arguing that the technical core of the job has not changed much and that patient-facing work is a smaller share than Amodei implies. The American College of Radiology's June 10, 2026 bulletin piece ["Radiologists Will Adapt"](https://www.acr.org/Clinical-Resources/Publications-and-Research/ACR-Bulletin/2026/radiologists-will-adapt) cites projections that imaging demand will grow at least as fast as the radiology workforce, which describes a shortage.

Amodei offers no study for the scan-reading claim in the episode, so treat it as his assertion.

## Bottom line

Amodei did say coding goes away first, and he also said end-to-end software engineering will follow, with the remaining human share getting smaller near 99%. His 20x figure is arithmetic under ideal conditions, and the one controlled trial measured a 19% slowdown with early-2025 tools. The radiologist comparison he leans on is contested by radiologists and by the ACR's workforce data.

## Sources

- [The AI Tsunami is Here & Society Isn't Ready](https://www.youtube.com/watch?v=68ylaeBbdsg) - the full People by WTF episode, February 24, 2026
- [Episode transcript](https://singjupost.com/people-by-wtf-w-dario-amodei-the-ai-tsunami-is-here-transcript/) - the clip's lines verbatim
- [CXOToday episode writeup](https://cxotoday.com/media-coverage/nikhil-kamath-and-anthropic-ceo-dario-amodei-discuss-ais-next-frontier-biotech-acceleration-and-the-evolving-role-of-data/) - publication date and guests
- [Anthropic's India expansion announcement](https://www.anthropic.com/news/expanding-global-operations-to-india) - Amodei's title and the Bengaluru office
- [METR randomized trial](https://metr.org/blog/2025-07-10-early-2025-ai-experienced-os-dev-study/) - measured vs self-reported AI speedup
- [Ben White, "Dario Dreams of Electric Radiology"](https://www.benwhite.com/radiology/dario-dreams-of-electric-radiology/) - a radiologist's rebuttal
- [ACR Bulletin, "Radiologists Will Adapt"](https://www.acr.org/Clinical-Resources/Publications-and-Research/ACR-Bulletin/2026/radiologists-will-adapt) - radiology workforce projections
- [Entrepreneur on Amodei at Davos](https://www.entrepreneur.com/business-news/ai-ceo-says-software-engineers-could-be-replaced-in-months/502087) - his six-to-twelve-month timeline


## Video description and verification notes

**The file is a 129.99-second MP4 (QuickTime/MOV container, major brand `isom`) at 576x1024, 9:16, 30 fps, H.264 High profile level 3.1, yuv420p, bt709, 3,899 video frames at 894,733 bps, paired with an HE-AACv2 stereo track at 44,100 Hz and 32,045 bps, for an overall 933,284 bps and 15,164,120 bytes on disk. It was posted 2026-10-07 at 20:52:47 UTC by handle `investbesrclips` (uploader ID 7689770681483445270), whose channel nickname is "Invest Best Clips." The audio is credited to "original sound" by Invest Best Clips, which is the standard TikTok credit for a creator's own upload rather than a licensed track. The description runs 88 characters and repeats the title plus five hashtags. At capture on 2026-10-08 the post had 306 views, 8 likes, 4 saves, 0 comments, and 0 reposts. I read 15 of the 65 extracted frames (001, 002, 003, 005, 010, 013, 015, 022, 030, 035, 040, 047, 051, 058, 065) and all 387 words of the transcript. The setting is a high-floor hotel suite with floor-to-ceiling windows looking out on tree canopy and a hazy city skyline, beige curtains, cream upholstered armchairs, a grey rug with a gold grid pattern, and a round gold side table holding a plant. Two men sit facing each other wearing lavalier mics. One wears a rust linen shirt, light jeans, a steel sports watch and beaded bracelets, and holds a tablet in a folio case. The other wears a navy ribbed short-sleeve polo, dark trousers, navy running shoes, round blue-framed glasses, a steel watch and a wedding band. For roughly the first four seconds a white banner spans the top of the frame reading "Anthropic CEO: Coding is dying / Learn THIS instead 👀" with an eyes emoji. The banner then disappears for the remaining two minutes. Throughout, bold condensed white all-caps karaoke captions with a dark outline flash one word at a time near the center: "CODING," "INDUSTRY," "WHAT," "HAS," "MEAN," "BACK," "20," "FOCUSED," "LIKE," "AMPLIFIED," "SAY," "AT," "THINK," "RADIOLOGIST." There is no lower-third name key, no podcast logo, no episode title, and no source attribution anywhere in the frame.**

## The speaker is Dario Amodei and the source is a February 2026 episode the clip never names

The clip identifies its guest only as "Anthropic CEO" in a banner that vanishes after four seconds. That label is correct. The speaker is Dario Amodei, co-founder and CEO of Anthropic, a title Anthropic itself uses in its [India expansion announcement](https://www.anthropic.com/news/expanding-global-operations-to-india) from October 7, 2025. The interviewer is Nikhil Kamath, the Zerodha co-founder who hosts the People by WTF podcast, which matches the clip's line about "my audience, which is the 20 or 25-year-old boy or girl in India."

The footage comes from "The AI Tsunami is Here & Society Isn't Ready," published February 24, 2026 and running about 69 minutes ([full episode](https://www.youtube.com/watch?v=68ylaeBbdsg), [coverage](https://cxotoday.com/media-coverage/nikhil-kamath-and-anthropic-ceo-dario-amodei-discuss-ais-next-frontier-biotech-acceleration-and-the-evolving-role-of-data/)). A [published transcript](https://singjupost.com/people-by-wtf-w-dario-amodei-the-ai-tsunami-is-here-transcript/) of that episode contains the clip's lines verbatim and places the recording in Bangalore. Anthropic's Bengaluru office opened in early 2026, and Amodei was in India that February. So this TikTok posted on October 7, 2026 recycles an exchange that is roughly seven and a half months old at upload. Nothing in the video says so.

## The clip opens with the answer and then rewinds to the question

The first line of the transcript is the punchline: "I think coding is going away first." The second line is Kamath starting his question from the beginning. The editor lifted Amodei's answer from the 24th line of the exchange and bolted it onto the front as a hook, then let the conversation play in order.

In the actual conversation, Kamath asks what industry gets disrupted and what has runway left, framed around a young person choosing a book, a college, or a skill. Amodei answers in general terms about human-centered work. Kamath then pushes back and asks him to separate coding from engineering, treating them as two different things. Only then does Amodei give the line the clip leads with. The distinction the quote depends on was Kamath's setup, not Amodei's framing. A viewer who stops after the first two seconds gets the verdict without the question that produced it.

## Amodei's full answer contradicts the banner's "learn THIS instead"

The banner promises a safe alternative. The audio under it does not deliver one. Amodei says coding goes first, "and then the broader task of software engineering will take longer." He then adds, in the same breath, "But I think that is, doing that end-to-end, I think that is going to happen as well." He names design, user understanding, demand sensing, and managing teams of AI models as things that "may still be present." The hedge is his own word, twice.

He closes the economic point with a ceiling rather than a promise: "at some point, you get to 99%, and then it becomes harder." That is the same person saying the refuge shrinks. The banner sells durability that the speaker declines to claim.

For calibration on his timelines, Amodei told The Economist at the World Economic Forum in January 2026 that AI models could do "most, maybe all" of what software engineers do within six to twelve months ([report](https://www.entrepreneur.com/business-news/ai-ceo-says-software-engineers-could-be-replaced-in-months/502087), published January 21, 2026). That window closed between January and roughly January 2027. Judge the February claim against that record rather than against the TikTok caption.

## The "20 times more productive" number is division, not measurement

Amodei's arithmetic is explicit in the clip. If a human does 5% of a task and AI does 95%, the human's output multiplies by 20. That is 1 divided by 0.05. It is an upper bound that holds only if the human's remaining 5% never becomes the bottleneck, and only if the AI's 95% needs no rework.

The one randomized controlled trial on this question points the other way in its setting. [METR](https://metr.org/blog/2025-07-10-early-2025-ai-experienced-os-dev-study/) had 16 experienced open-source developers work 246 real issues from their own repositories, with AI tool access randomly permitted or forbidden per issue. Developers forecast a 24% speedup and reported a 20% speedup afterward. Measured, they took 19% longer with AI access. METR states plainly that it does "not claim that our developers or repositories represent a majority or plurality of software development work," and that the result is a snapshot of early-2025 tools. So it does not refute Amodei. It does show that the gap between felt and measured speedup ran 39 percentage points wide in one careful study, which is reason to treat a 20x figure as a model rather than an observation.

## The radiologist example the clip ends on was never shown, and its premise is disputed

The last line of the clip is "similar to the radiologist example I gave." The clip never gives it. Earlier in the full episode Amodei argued that AI has gotten better than radiologists at reading scans, that the number of radiologists has not fallen, and that what remains is walking the patient through results.

Radiologists pushed back on exactly that framing. Ben White, a practicing radiologist, called the claim ["completely baseless and without a reality correlate"](https://www.benwhite.com/radiology/dario-dreams-of-electric-radiology/), arguing that the technical core of the job has not meaningfully changed and that the patient-facing share of the work is far smaller than Amodei implies. The American College of Radiology's June 10, 2026 bulletin piece ["Radiologists Will Adapt"](https://www.acr.org/Clinical-Resources/Publications-and-Research/ACR-Bulletin/2026/radiologists-will-adapt) cites work by its Commission on Human Resources and the Harvey Neiman Health Policy Institute projecting that imaging demand will grow at least as fast as the radiology workforce, which describes a shortage rather than a displacement.

Treat "AI is better than radiologists at reading scans" as Amodei's assertion rather than an established finding. He offers no study in the episode, and the workforce data cited by the specialty's own trade body points to growth.

## Key Takeaways

- **Verified:** The banner's "Anthropic CEO" label is accurate. The speaker is Dario Amodei, and the quotes match the published transcript of People by WTF, February 24, 2026, recorded in Bangalore.
- **Clarification:** The clip is a hook edit. Its opening line is Amodei's answer, pulled forward from later in the exchange. The coding-versus-engineering split was Nikhil Kamath's framing, offered as a follow-up question, not Amodei's own distinction.
- **Correction:** The banner "Coding is dying / Learn THIS instead" implies a durable alternative. In the same answer, Amodei says end-to-end software engineering "is going to happen as well" and that the comparative-advantage zone runs out near 99%. The clip's own audio refuses the promise its title card makes.
- **Partial correction:** "20 times more productive" is 1 divided by 0.05 under ideal conditions, not a measured result. METR's randomized trial of 16 developers on 246 tasks measured a 19% slowdown with early-2025 AI tools against a self-reported 20% speedup.
- **Unverified:** The radiologist premise the clip's last line points back to is Amodei's claim, not a cited finding. A practicing radiologist disputes it, and the ACR projects workforce growth.
- **Context:** The post is dated October 7, 2026 and carries no date, episode title, or link. The underlying conversation is from February 2026.
- **Context:** The auto-transcript renders the first word as "codeine" rather than "coding." That is a speech-recognition error in the pipeline, not anything said or shown in the video.

## Resources

- [The AI Tsunami is Here & Society Isn't Ready](https://www.youtube.com/watch?v=68ylaeBbdsg) is the full People by WTF episode the clip was cut from, published February 24, 2026.
- [People by WTF w/ Dario Amodei transcript](https://singjupost.com/people-by-wtf-w-dario-amodei-the-ai-tsunami-is-here-transcript/) carries the clip's lines verbatim and places the recording in Bangalore.
- [CXOToday's episode writeup](https://cxotoday.com/media-coverage/nikhil-kamath-and-anthropic-ceo-dario-amodei-discuss-ais-next-frontier-biotech-acceleration-and-the-evolving-role-of-data/) establishes the February 24, 2026 publication date and the Kamath-Amodei pairing.
- [Anthropic's India expansion announcement](https://www.anthropic.com/news/expanding-global-operations-to-india), October 7, 2025, establishes Amodei's title as CEO and co-founder and the Bengaluru office timing.
- [METR's randomized trial](https://metr.org/blog/2025-07-10-early-2025-ai-experienced-os-dev-study/) establishes the 19% measured slowdown against a 20% self-reported speedup, with 16 developers and 246 tasks.
- [Ben White, "Dario Dreams of Electric Radiology"](https://www.benwhite.com/radiology/dario-dreams-of-electric-radiology/) establishes a practicing radiologist's rebuttal to the scan-reading claim.
- [ACR Bulletin, "Radiologists Will Adapt"](https://www.acr.org/Clinical-Resources/Publications-and-Research/ACR-Bulletin/2026/radiologists-will-adapt), June 10, 2026, establishes the ACR and Neiman Institute projection that imaging demand grows at least as fast as the radiology workforce.
- [Entrepreneur's report on Amodei at Davos](https://www.entrepreneur.com/business-news/ai-ceo-says-software-engineers-could-be-replaced-in-months/502087), January 21, 2026, establishes his "most, maybe all" six-to-twelve-month software engineering timeline given to The Economist.
