Escape Meridian

MIT Technology Review · Jun 18, 2024

MIT Technology Review’s The Download examines AI’s limitations

Markdown · JSON · Original source

MIT Technology Review’s The Download was published on June 18, 2024. It calls AI’s tendency to make things up, known as , one of the biggest obstacles to wider chatbot adoption. The newsletter reports that Health Organization’s chatbot SARAH gave incorrect information, including fake clinic names and addresses, and recalls other chatbot failures. It directs readers to Will Douglas Heaven’s article, which asks why chatbots hallucinate and why the problem cannot be fixed. The edition also describes a study in which researchers asked professional comedians to use AI language models to write jokes and performances. The tools helped with initial drafts and structure, but the newsletter says AI did not produce work that was original, stimulating, or funny. It argues that AI can take on formulaic parts of creative work but cannot replace the magic and originality humans bring. The newsletter does not explain why chatbots hallucinate or whether the problem can be fixed.

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.

MIT Technology ReviewArticle · technologyreview.comOpen original source

Original source

MIT Technology Review · Article

Source details

Original source date: Jun 18, 2024 (2024-06-18T12:10:00Z).

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

Added to Escape Meridian: 2026-11-18T12:23:11.798Z. First observed: 2026-10-29T02:47:56.227Z. Retrieved: 2026-10-29T02:49:34.567Z.

Referenced by

Recorded citations in public readings. A reference alone does not establish agreement or influence.

No public references recorded here yet.

Reference data (JSON)

Limits

The edition points readers to the full article on chatbot ; that article’s explanation of causes and possible fixes is not included in the retrieved text. The newsletter gives no further detail about the comedians’ study methods or how its results were assessed. Text as retrieved on October 29, 2026; originally published June 18, 2024.

Open question

Why do chatbots , and why can’t we fix it?

Source-stated

The Download describes hallucination as a major obstacle to wider chatbot adoption.

This tendency to make things up—known as hallucination—is one of the biggest obstacles holding chatbots back from more widespread adoption.

Source-stated

The newsletter reports that SARAH gave incorrect information, including fake clinic names and addresses.

In one case, it came up with a list of fake names and addresses for nonexistent clinics in San Francisco.

Source-stated

The newsletter says AI tools helped comedians with drafts and structure but did not produce work the study characterized as original, stimulating, or funny.

Although the tools helped them to produce initial drafts and structure their routines, AI was not able to produce anything that was original, stimulating, or, crucially, funny.

Source-stated

The Download argues that AI can handle some formulaic creative tasks but cannot replace the originality humans bring.

It can take on some of the boring, mundane, formulaic aspects of the creative process, but it can’t replace the magic and originality that humans bring.

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.

Hallucination

When an AI system states something false or made up as if it were true.

The World

The World is a public radio news program publishing international reporting on politics, society, labor and technology.

Google DeepMind

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

Editorial provenance

Version 1. Publication ID: 4fb4691c-c236-be64-34cb-d0932fa4e28e.

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

Source check: passed by anthropic/claude-opus-5.5 on 2026-11-18T12:20:03.140Z. 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: db3ec12bc8b717c8c47313055b03e8a8f177350bc7028a3681a464976e20674e. Source text hash: 71a49be34bbfe9016a99464d7cf01f51ec66fac26446f1a6f2d77a63dd13e8e3.

Revision and source-check history

Version 1 · 2026-11-18T12:23:11.798Z

MIT Technology Review’s The Download examines AI’s limitations

MIT Technology Review’s The Download was published on June 18, 2024. It calls AI’s tendency to make things up, known as , one of the biggest obstacles to wider chatbot adoption. The newsletter reports that Health Organization’s chatbot SARAH gave incorrect information, including fake clinic names and addresses, and recalls other chatbot failures. It directs readers to Will Douglas Heaven’s article, which asks why chatbots hallucinate and why the problem cannot be fixed. The edition also describes a study in which researchers asked professional comedians to use AI language models to write jokes and performances. The tools helped with initial drafts and structure, but the newsletter says AI did not produce work that was original, stimulating, or funny. It argues that AI can take on formulaic parts of creative work but cannot replace the magic and originality humans bring. The newsletter does not explain why chatbots hallucinate or whether the problem can be fixed.

Original source

MIT Technology Review · Article

Source details

Original source date: Jun 18, 2024 (2024-06-18T12:10:00Z).

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

Added to Escape Meridian: 2026-11-18T12:23:11.798Z. First observed: 2026-10-29T02:47:56.227Z. Retrieved: 2026-10-29T02:49:34.567Z.

Limits

The edition points readers to the full article on chatbot ; that article’s explanation of causes and possible fixes is not included in the retrieved text. The newsletter gives no further detail about the comedians’ study methods or how its results were assessed. Text as retrieved on October 29, 2026; originally published June 18, 2024.

Open question

Why do chatbots , and why can’t we fix it?

Source-stated

The Download describes hallucination as a major obstacle to wider chatbot adoption.

This tendency to make things up—known as hallucination—is one of the biggest obstacles holding chatbots back from more widespread adoption.

Source-stated

The newsletter reports that SARAH gave incorrect information, including fake clinic names and addresses.

In one case, it came up with a list of fake names and addresses for nonexistent clinics in San Francisco.

Source-stated

The newsletter says AI tools helped comedians with drafts and structure but did not produce work the study characterized as original, stimulating, or funny.

Although the tools helped them to produce initial drafts and structure their routines, AI was not able to produce anything that was original, stimulating, or, crucially, funny.

Source-stated

The Download argues that AI can handle some formulaic creative tasks but cannot replace the originality humans bring.

It can take on some of the boring, mundane, formulaic aspects of the creative process, but it can’t replace the magic and originality that humans bring.

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.

Hallucination

When an AI system states something false or made up as if it were true.

The World

The World is a public radio news program publishing international reporting on politics, society, labor and technology.

Google DeepMind

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

Editorial provenance

Version 1. Publication ID: 4fb4691c-c236-be64-34cb-d0932fa4e28e.

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

Source check: passed by anthropic/claude-opus-5.5 on 2026-11-18T12:20:03.140Z. 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: db3ec12bc8b717c8c47313055b03e8a8f177350bc7028a3681a464976e20674e. Source text hash: 71a49be34bbfe9016a99464d7cf01f51ec66fac26446f1a6f2d77a63dd13e8e3.

passed · 2026-11-18T12:20:03.140Z · anthropic/claude-opus-5.5

Publication: 4fb4691c-c236-be64-34cb-d0932fa4e28e

attribution — pass: This is one newsletter issue that comments on its stories in its own words, not a bare listing. The reading attributes claims to MIT Technology Review's The Download and credits Will Douglas Heaven's article. The comedy segment, originally by Melissa Heikkilä from The Algorithm, is described as what the newsletter says or argues. That is acceptable because it appears in this edition.

This tendency to make things up—known as hallucination—is one of the biggest obstacles holding chatbots back from more widespread adoption.

summary — pass: The reading accurately conveys the two AI segments: hallucination as an obstacle, with the WHO SARAH example, and the DeepMind comedian study with the newsletter's argument about creative limits. Hallucination is defined beside the reading. The reading correctly says the newsletter does not explain the causes of hallucination. The question is taken from the source and remains open.

Why do they do it? And why can’t we fix it? Read the full story.

date — pass: The June 18, 2024 date matches the UTC day of originalPublished. The Deep Dive sidebar lists later stories, such as OpenAI's math findings, but the reading neither uses nor relies on them. No forecasts are presented as outcomes. The kind is publisher-page, so no archive disclosure is needed. The modification timestamp precedes publication.

support — pass: Each claim matches its quoted source text. The summary's statements about SARAH, the comedians study and the creativity argument are source-stated and attributed. No benefits, costs or forecasts are invented.

It can take on some of the boring, mundane, formulaic aspects of the creative process, but it can’t replace the magic and originality that humans bring.

media — pass: No media was selected and none is available among the verified candidates, so there is no media identity or relevance issue.

Challenge or correct this reading

Send a correction or an opposing source with this reading’s URL. Disagreement and uncertainty can remain unresolved.