# MIT Technology Review reports on comedians using AI to write routines

[Canonical reading](https://escapemeridian.com/readings/mit-technology-review-reports-on-comedians-using-ai-to-write-360ecdf9)

MIT Technology Review published this report on June 17, 2024. Google DeepMind researchers led by Piotr Mirowski studied professional comedians who already used AI in their creative work. The comedians said AI could help structure a monologue or produce a rough draft, but its jokes often felt bland and generic, and they did not feel particularly proud of the results. Some found it useful for getting past a blank page; others felt the effort was not worth the reward.

The report connects those reactions to safety filters that can hinder offensive or sexually suggestive jokes and dark humor, and to models’ reliance on patterns in their training data. Participants also said a model would not generate comedy from the perspective of an Asian woman, though it would from the perspective of a white man; they felt this reinforced the status quo and erased minority perspectives. Computer science researcher Tuhin Chakrabarty argues that comedy depends on surprise and deviation from the norm, while language models predict one word at a time. He expects people to use these tools for scripts, screenplays, and advertisements, but argues that creative and comedic writing depends on experience and “vibes,” not an algorithm. The report describes the comedians’ experiences, not a measure of how well AI-written comedy performs with audiences.

Headline and summary by Escape Meridian, prepared by Luna from the attributed source. A source check is not independent confirmation of the source’s claims.

## Original source

MIT Technology Review · Article

Original source date: Jun 17, 2024 (2024-06-17T08:00:00Z).

Date basis: Date from the publisher’s own records, through its API.

Added to Escape Meridian: 2026-11-18T12:23:10.245Z. First observed: 2026-10-28T23:57:52.462Z. Retrieved: 2026-10-28T23:59:49.995Z.

- [What happened when 20 comedians got AI to write their routines](https://www.technologyreview.com/2024/06/17/1093740/what-happened-when-20-comedians-got-ai-to-write-their-routines/)

## Limits

Text as retrieved on October 28, 2026; originally published June 17, 2024. The report describes comedians’ experiences and views; it gives no measure of audience response to AI-written comedy.


## Open question

How would audiences respond to comedy written with AI, and would their reactions differ across kinds of jokes or perspectives?


## Source-stated

Google DeepMind researchers led by Piotr Mirowski studied professional comedians who already used AI in their artistic process.

> Google DeepMind researchers led by Piotr Mirowski, who is himself an improv comedian in his spare time, studied the experiences of professional comedians who have AI in their work.

- [Supporting source](https://www.technologyreview.com/2024/06/17/1093740/what-happened-when-20-comedians-got-ai-to-write-their-routines/)

## Source-stated

The comedians reported enjoying AI use but not feeling particularly proud of the resulting material.

> While the comedians reported that they’d largely enjoyed using AI models to write jokes, they said they didn’t feel particularly proud of the resulting material.

- [Supporting source](https://www.technologyreview.com/2024/06/17/1093740/what-happened-when-20-comedians-got-ai-to-write-their-routines/)

## Source-stated

Some comedians found AI useful for starting a draft, while others thought the effort was not worth the reward.

> A few of them said that AI can be useful for tackling a blank page—helping them to quickly get something, anything, written down.

- [Supporting source](https://www.technologyreview.com/2024/06/17/1093740/what-happened-when-20-comedians-got-ai-to-write-their-routines/)

## Source-stated

Participants said a model would not generate comedy from the perspective of an Asian woman, though it could from the perspective of a white man.

> Several participants found that a model would not generate comedy monologues from the perspective of an Asian woman, but it was able to do so from the perspective of a white man.

- [Supporting source](https://www.technologyreview.com/2024/06/17/1093740/what-happened-when-20-comedians-got-ai-to-write-their-routines/)

## Source-stated

Tuhin Chakrabarty argues that comedy relies on surprise and incongruity, which conflict with how these models work.

> So much of humor relies on being surprising and incongruous, which is at odds with how these models work, says Tuhin Chakrabarty

- [Supporting source](https://www.technologyreview.com/2024/06/17/1093740/what-happened-when-20-comedians-got-ai-to-write-their-routines/)

## Source-stated

Chakrabarty expects people to use AI tools for scripts, screenplays, and advertisements, but says creative and comedic writing depends on experience and vibes.

> “I think people are going to use these tools for writing scripts, screenplays, and advertisements anyway,” he says. “But true creative and comedic writing is based on experience and vibes. Not an algorithm.”

- [Supporting source](https://www.technologyreview.com/2024/06/17/1093740/what-happened-when-20-comedians-got-ai-to-write-their-routines/)

## Source and contributor context

MIT Technology Review · Publisher

The publisher is identified here. No individual author has been inferred.


## Source media

Media retrieval status: unchecked.


## Terms and references

Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.


## Google DeepMind

Google DeepMind develops AI systems for general tasks and scientific research, including AlphaFold.

- [Explore this context](/entities/google-deepmind)
- [Reference](https://deepmind.google/blog/)
- [Reference](https://deepmind.google/science/alphafold/)

## TIME

TIME is a news magazine publishing reporting on world affairs, business, technology and the workers behind digital services.

- [Explore this context](/entities/time)
- [Reference](https://time.com/)

## Editorial provenance

Version 1. Publication ID: 695ba978-6dcc-cf73-d2c0-5866e564af6f.

Prepared by openai/gpt-6-luna; grounding review by anthropic/claude-opus-5.5.

Source check: passed by anthropic/claude-opus-5.5 on 2026-11-18T12:21:01.274Z. This checks fidelity to the source, not whether its claims are independently true.

Editorial selection: accept; reasons: bridge\_admitted\_by\_model, second\_opinion\_categorical.

Observation: 08c023a7c0c1eb9fc709ba77b676d425cc9bfa48b8efccc9063972a52d9f6e88. Source text hash: fc56f8bfc6e9771f77eadab8e972c55ae5e37b14b6f9728cf55ce247cc78a803.


## Revision and source-check history

### Version 1 — 2026-11-18T12:23:10.245Z

MIT Technology Review published this report on June 17, 2024. Google DeepMind researchers led by Piotr Mirowski studied professional comedians who already used AI in their creative work. The comedians said AI could help structure a monologue or produce a rough draft, but its jokes often felt bland and generic, and they did not feel particularly proud of the results. Some found it useful for getting past a blank page; others felt the effort was not worth the reward.

The report connects those reactions to safety filters that can hinder offensive or sexually suggestive jokes and dark humor, and to models’ reliance on patterns in their training data. Participants also said a model would not generate comedy from the perspective of an Asian woman, though it would from the perspective of a white man; they felt this reinforced the status quo and erased minority perspectives. Computer science researcher Tuhin Chakrabarty argues that comedy depends on surprise and deviation from the norm, while language models predict one word at a time. He expects people to use these tools for scripts, screenplays, and advertisements, but argues that creative and comedic writing depends on experience and “vibes,” not an algorithm. The report describes the comedians’ experiences, not a measure of how well AI-written comedy performs with audiences.

## Original source

MIT Technology Review · Article

Original source date: Jun 17, 2024 (2024-06-17T08:00:00Z).

Date basis: Date from the publisher’s own records, through its API.

Added to Escape Meridian: 2026-11-18T12:23:10.245Z. First observed: 2026-10-28T23:57:52.462Z. Retrieved: 2026-10-28T23:59:49.995Z.

- [What happened when 20 comedians got AI to write their routines](https://www.technologyreview.com/2024/06/17/1093740/what-happened-when-20-comedians-got-ai-to-write-their-routines/)

## Limits

Text as retrieved on October 28, 2026; originally published June 17, 2024. The report describes comedians’ experiences and views; it gives no measure of audience response to AI-written comedy.


## Open question

How would audiences respond to comedy written with AI, and would their reactions differ across kinds of jokes or perspectives?


## Source-stated

Google DeepMind researchers led by Piotr Mirowski studied professional comedians who already used AI in their artistic process.

> Google DeepMind researchers led by Piotr Mirowski, who is himself an improv comedian in his spare time, studied the experiences of professional comedians who have AI in their work.

- [Supporting source](https://www.technologyreview.com/2024/06/17/1093740/what-happened-when-20-comedians-got-ai-to-write-their-routines/)

## Source-stated

The comedians reported enjoying AI use but not feeling particularly proud of the resulting material.

> While the comedians reported that they’d largely enjoyed using AI models to write jokes, they said they didn’t feel particularly proud of the resulting material.

- [Supporting source](https://www.technologyreview.com/2024/06/17/1093740/what-happened-when-20-comedians-got-ai-to-write-their-routines/)

## Source-stated

Some comedians found AI useful for starting a draft, while others thought the effort was not worth the reward.

> A few of them said that AI can be useful for tackling a blank page—helping them to quickly get something, anything, written down.

- [Supporting source](https://www.technologyreview.com/2024/06/17/1093740/what-happened-when-20-comedians-got-ai-to-write-their-routines/)

## Source-stated

Participants said a model would not generate comedy from the perspective of an Asian woman, though it could from the perspective of a white man.

> Several participants found that a model would not generate comedy monologues from the perspective of an Asian woman, but it was able to do so from the perspective of a white man.

- [Supporting source](https://www.technologyreview.com/2024/06/17/1093740/what-happened-when-20-comedians-got-ai-to-write-their-routines/)

## Source-stated

Tuhin Chakrabarty argues that comedy relies on surprise and incongruity, which conflict with how these models work.

> So much of humor relies on being surprising and incongruous, which is at odds with how these models work, says Tuhin Chakrabarty

- [Supporting source](https://www.technologyreview.com/2024/06/17/1093740/what-happened-when-20-comedians-got-ai-to-write-their-routines/)

## Source-stated

Chakrabarty expects people to use AI tools for scripts, screenplays, and advertisements, but says creative and comedic writing depends on experience and vibes.

> “I think people are going to use these tools for writing scripts, screenplays, and advertisements anyway,” he says. “But true creative and comedic writing is based on experience and vibes. Not an algorithm.”

- [Supporting source](https://www.technologyreview.com/2024/06/17/1093740/what-happened-when-20-comedians-got-ai-to-write-their-routines/)

## Source and contributor context

MIT Technology Review · Publisher

The publisher is identified here. No individual author has been inferred.


## Source media

Media retrieval status: unchecked.


## Terms and references

Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.


## Google DeepMind

Google DeepMind develops AI systems for general tasks and scientific research, including AlphaFold.

- [Explore this context](/entities/google-deepmind)
- [Reference](https://deepmind.google/blog/)
- [Reference](https://deepmind.google/science/alphafold/)

## TIME

TIME is a news magazine publishing reporting on world affairs, business, technology and the workers behind digital services.

- [Explore this context](/entities/time)
- [Reference](https://time.com/)

## Editorial provenance

Version 1. Publication ID: 695ba978-6dcc-cf73-d2c0-5866e564af6f.

Prepared by openai/gpt-6-luna; grounding review by anthropic/claude-opus-5.5.

Source check: passed by anthropic/claude-opus-5.5 on 2026-11-18T12:21:01.274Z. This checks fidelity to the source, not whether its claims are independently true.

Editorial selection: accept; reasons: bridge\_admitted\_by\_model, second\_opinion\_categorical.

Observation: 08c023a7c0c1eb9fc709ba77b676d425cc9bfa48b8efccc9063972a52d9f6e88. Source text hash: fc56f8bfc6e9771f77eadab8e972c55ae5e37b14b6f9728cf55ce247cc78a803.


### passed — 2026-11-18T12:21:01.274Z — anthropic/claude-opus-5.5

Publication: 695ba978-6dcc-cf73-d2c0-5866e564af6f

attribution — pass: The reading attributes the report to MIT Technology Review, the study to Google DeepMind researchers led by Piotr Mirowski, the comedians' views to participants, and the arguments about surprise and 'vibes' to Tuhin Chakrabarty. The source is a single article, not a listing.

> Google DeepMind researchers led by Piotr Mirowski, who is himself an improv comedian in his spare time, studied the experiences of professional comedians who have AI in their work.

summary — pass: The reading leads with the actor and the finding: AI helps with structure and rough drafts, but its jokes feel bland. It explains safety filters, the bias finding and Chakrabarty's argument in plain language, and it states the study's scope. Omitting the cruise-ship quote and Colleen Lavin's view softens some of the source's voice but loses nothing essential. The question about audience response follows from the stated gap.

> they struggled to produce material that was original, stimulating, or—crucially—funny

date — pass: The originalPublished instant, 2024-06-17, matches the reading's June 17, 2024 date. The page was modified in December 2024, but every passage the reading uses is plainly part of the original article body. The later Deep Dive sidebar items, such as the AlphaGo ten-year piece, are not used. There is no hindsight framing, and this is a publisher page, not an archive copy.

> They presented their findings at the ACM FAccT conference in Rio earlier this month

support — pass: Each claim matches a source passage: enjoyment without pride, blank-page help versus effort not worth it, the Asian woman/white man refusal, safety filters, and Chakrabarty's arguments. The limitation that there is no audience measure fairly describes a study of the comedians' own experiences. Although the page was edited after publication, the cited text is plainly part of the original article body.

> Others felt that the amount of effort just wasn’t worth the reward.

media — pass: No media was selected, and no verified candidates were supplied, so null is appropriate.

## Recorded references

Explicit citations in public readings to this reading. Shared topics are not citations. Other record types need an explicit reference adapter.

No public references recorded here yet.

[Reference data (JSON)](/readings/360ecdf91e9ed72a2da92f3238f376fa57dbc3ccff941a3817a96b11e913b0cc/references.json)

## Corrections and challenges

[Send a correction or an opposing source](https://escapemeridian.com/corrections/)

[JSON record](https://escapemeridian.com/readings/mit-technology-review-reports-on-comedians-using-ai-to-write-360ecdf9.json)
