# MIT Technology Review reports on chatbots, fact-checking AI and building design

[Canonical reading](https://escapemeridian.com/readings/mit-technology-review-reports-on-chatbots-fact-checking-ai-and-b48cb986)

MIT Technology Review’s September 13, 2024, edition of The Download reports on research into chatbot conversations about conspiracy theories, Google’s DataGemma tool, and a laboratory where neuroscientists and architects study building design. Researchers from MIT Sloan and Cornell University found that chatting with a large language model reduced participants’ belief in a conspiracy theory by about 20%, including among participants who said those beliefs were important to their identity. The newsletter presents this as a possible way to engage and educate people who hold baseless theories; it does not give the full study’s methods or results here. It also reports that Google is releasing DataGemma to help large language models check responses against reliable data and cite sources. Google hopes this will reduce incorrect claims, but the newsletter says the tool comes with caveats and describes its possible benefit to Google’s search plans conditionally. A third item describes neuroscientists and architects working together in an East London laboratory with simulated worlds, addressing how building designs can suit people with different minds, bodies and needs. The newsletter’s accounts introduce these efforts, but do not establish how well the chatbot approach or DataGemma works in practice, or what the laboratory’s work has produced.

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: Sep 13, 2024 (2024-09-13T12:10:00Z).

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

Added to Escape Meridian: 2026-11-17T06:13:23.221Z. First observed: 2026-10-29T22:50:18.622Z. Retrieved: 2026-10-29T22:51:46.040Z.

- [The Download: conspiracy-debunking chatbots, and fact-checking AI](https://www.technologyreview.com/2024/09/13/1103952/the-download-conspiracy-debunking-chatbots-and-fact-checking-ai/)

## Limits

Text as retrieved on October 29, 2026; originally published September 13, 2024. The newsletter points readers to full stories, but the retrieved text does not include the chatbot study’s methods or further results, details of DataGemma’s caveats, or outcomes from the laboratory’s work. No media was supplied.


## Open question

How durable is the reduction in conspiracy-theory belief after chatbot conversations, and what caveats affect DataGemma’s fact-checking results?


## Source-stated

Researchers from MIT Sloan and Cornell University found that chatting about a conspiracy theory with a large language model reduced participants’ belief in it by about 20%, including among people who said the belief was important to their identity.

> Researchers from MIT Sloan and Cornell University found that chatting about a conspiracy theory with a large language model (LLM) reduced people’s belief in it by about 20%—even among participants who claimed that their beliefs were important to their identity

- [Supporting source](https://www.technologyreview.com/2024/09/13/1103952/the-download-conspiracy-debunking-chatbots-and-fact-checking-ai/)

## Source-stated

MIT Technology Review reports that Google is releasing DataGemma to help large language models check responses against reliable data and cite sources more transparently.

> Google is releasing a tool called DataGemma that it hopes will help to reduce problems caused by AI 'hallucinating’, or making incorrect claims. It uses two methods to help large language models fact-check their responses against reliable data and cite their sources more transparently to users.

- [Supporting source](https://www.technologyreview.com/2024/09/13/1103952/the-download-conspiracy-debunking-chatbots-and-fact-checking-ai/)

## Source-stated

The newsletter says DataGemma could benefit Google’s plan to embed AI deeper into its search engine if it works as hoped, while noting that it comes with caveats.

> If it works as hoped, it could be a real boon for Google’s plan to embed AI deeper into its search engine. But it comes with a host of caveats.

- [Supporting source](https://www.technologyreview.com/2024/09/13/1103952/the-download-conspiracy-debunking-chatbots-and-fact-checking-ai/)

## Source-stated

The newsletter describes neuroscientists and architects working together at an East London laboratory that lets researchers build simulated worlds, with the aim of considering varied users’ needs in building design.

> To answer that question, neuroscientists and architects are joining forces at an enormous laboratory in East London—one that allows researchers to build simulated worlds.

- [Supporting source](https://www.technologyreview.com/2024/09/13/1103952/the-download-conspiracy-debunking-chatbots-and-fact-checking-ai/)

## 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.


## Large language model

A language model with many learned numerical settings, used to process and generate language.

- [Definition reference](https://developers.google.com/machine-learning/glossary#large-language-model)

## Editorial provenance

Version 1. Publication ID: 79160f2d-11f5-51b8-96e0-afcace220e64.

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

Source check: passed by anthropic/claude-opus-5.5 on 2026-11-17T06:08:39.231Z. 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: dfc6a1a98572dfe922c324ef69fa4adc72b0faa34c9db6ca5b5d72709cdfa8fe. Source text hash: 8e7db37af14d1fee060be470091e2857334707a2ac80c5068574240a5eeed7ad.


## Revision and source-check history

### Version 1 — 2026-11-17T06:13:23.221Z

MIT Technology Review’s September 13, 2024, edition of The Download reports on research into chatbot conversations about conspiracy theories, Google’s DataGemma tool, and a laboratory where neuroscientists and architects study building design. Researchers from MIT Sloan and Cornell University found that chatting with a large language model reduced participants’ belief in a conspiracy theory by about 20%, including among participants who said those beliefs were important to their identity. The newsletter presents this as a possible way to engage and educate people who hold baseless theories; it does not give the full study’s methods or results here. It also reports that Google is releasing DataGemma to help large language models check responses against reliable data and cite sources. Google hopes this will reduce incorrect claims, but the newsletter says the tool comes with caveats and describes its possible benefit to Google’s search plans conditionally. A third item describes neuroscientists and architects working together in an East London laboratory with simulated worlds, addressing how building designs can suit people with different minds, bodies and needs. The newsletter’s accounts introduce these efforts, but do not establish how well the chatbot approach or DataGemma works in practice, or what the laboratory’s work has produced.

## Original source

MIT Technology Review · Article

Original source date: Sep 13, 2024 (2024-09-13T12:10:00Z).

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

Added to Escape Meridian: 2026-11-17T06:13:23.221Z. First observed: 2026-10-29T22:50:18.622Z. Retrieved: 2026-10-29T22:51:46.040Z.

- [The Download: conspiracy-debunking chatbots, and fact-checking AI](https://www.technologyreview.com/2024/09/13/1103952/the-download-conspiracy-debunking-chatbots-and-fact-checking-ai/)

## Limits

Text as retrieved on October 29, 2026; originally published September 13, 2024. The newsletter points readers to full stories, but the retrieved text does not include the chatbot study’s methods or further results, details of DataGemma’s caveats, or outcomes from the laboratory’s work. No media was supplied.


## Open question

How durable is the reduction in conspiracy-theory belief after chatbot conversations, and what caveats affect DataGemma’s fact-checking results?


## Source-stated

Researchers from MIT Sloan and Cornell University found that chatting about a conspiracy theory with a large language model reduced participants’ belief in it by about 20%, including among people who said the belief was important to their identity.

> Researchers from MIT Sloan and Cornell University found that chatting about a conspiracy theory with a large language model (LLM) reduced people’s belief in it by about 20%—even among participants who claimed that their beliefs were important to their identity

- [Supporting source](https://www.technologyreview.com/2024/09/13/1103952/the-download-conspiracy-debunking-chatbots-and-fact-checking-ai/)

## Source-stated

MIT Technology Review reports that Google is releasing DataGemma to help large language models check responses against reliable data and cite sources more transparently.

> Google is releasing a tool called DataGemma that it hopes will help to reduce problems caused by AI 'hallucinating’, or making incorrect claims. It uses two methods to help large language models fact-check their responses against reliable data and cite their sources more transparently to users.

- [Supporting source](https://www.technologyreview.com/2024/09/13/1103952/the-download-conspiracy-debunking-chatbots-and-fact-checking-ai/)

## Source-stated

The newsletter says DataGemma could benefit Google’s plan to embed AI deeper into its search engine if it works as hoped, while noting that it comes with caveats.

> If it works as hoped, it could be a real boon for Google’s plan to embed AI deeper into its search engine. But it comes with a host of caveats.

- [Supporting source](https://www.technologyreview.com/2024/09/13/1103952/the-download-conspiracy-debunking-chatbots-and-fact-checking-ai/)

## Source-stated

The newsletter describes neuroscientists and architects working together at an East London laboratory that lets researchers build simulated worlds, with the aim of considering varied users’ needs in building design.

> To answer that question, neuroscientists and architects are joining forces at an enormous laboratory in East London—one that allows researchers to build simulated worlds.

- [Supporting source](https://www.technologyreview.com/2024/09/13/1103952/the-download-conspiracy-debunking-chatbots-and-fact-checking-ai/)

## 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.


## Large language model

A language model with many learned numerical settings, used to process and generate language.

- [Definition reference](https://developers.google.com/machine-learning/glossary#large-language-model)

## Editorial provenance

Version 1. Publication ID: 79160f2d-11f5-51b8-96e0-afcace220e64.

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

Source check: passed by anthropic/claude-opus-5.5 on 2026-11-17T06:08:39.231Z. 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: dfc6a1a98572dfe922c324ef69fa4adc72b0faa34c9db6ca5b5d72709cdfa8fe. Source text hash: 8e7db37af14d1fee060be470091e2857334707a2ac80c5068574240a5eeed7ad.


### passed — 2026-11-17T06:08:39.231Z — anthropic/claude-opus-5.5

Publication: 79160f2d-11f5-51b8-96e0-afcace220e64

attribution — pass: The source is one issue of MIT Technology Review's The Download newsletter, which summarizes stories in its own words, so it counts as one piece. The reading attributes the items to that newsletter and keeps the findings attributed to the MIT Sloan and Cornell researchers and to Google's hopes.

> This is today's edition of The Download, our weekday newsletter that provides a daily dose of what's going on in the world of technology.

summary — pass: The reading accurately summarizes the newsletter's three lead items and keeps the conditional framing of DataGemma's benefit and its caveats. Large language model is defined beside the reading. Hallucination is rendered as incorrect claims, and DataGemma is described in plain terms. The question joins two related unknowns: how lasting the belief reduction is, and DataGemma's caveats. That is a minor stylistic point, not a misreading.

> The findings could represent an important step forward in how we engage with and educate people who espouse baseless theories.

date — pass: September 13, 2024 matches the UTC day of originalPublished, 2024-09-13T12:10Z. The Deep Dive list contains apparently later items, such as 377 OpenAI math findings, but the reading neither uses nor relies on them. The evidence is a publisher page, not an archive copy, so no archive disclosure is needed. There is no hindsight framing.

support — pass: Each claim matches the source text: the roughly 20% belief reduction, including among identity-linked believers; DataGemma's two methods for fact-checking against reliable data and citing sources; the conditional benefit to Google's search plans with caveats; and the East London simulated-worlds laboratory. The stated limitations do not overreach.

> If it works as hoped, it could be a real boon for Google’s plan to embed AI deeper into its search engine. But it comes with a host of caveats.

media — pass: No media candidates were supplied and none was selected. The reading notes that no media was supplied.

## 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/b48cb986e93f7a04700cb263d04509b8a84b5c246c80a698a55a6d537f49535b/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-chatbots-fact-checking-ai-and-b48cb986.json)
