MIT Technology Review · Mar 11, 2025
MIT Technology Review’s The Download on AI fairness and AGI hype
Markdown · JSON · Original source
MIT Technology Review’s March 11, 2025, edition of The Download says new AI could help developers judge fairness, while actual fixes may require other techniques. Scott J Mulligan reports that the benchmarks assess AI systems’ awareness of scenarios and contexts; he says their researchers were prompted by earlier approaches whose failure to account for differences between groups could make systems less fair. The edition does not name the benchmarks or describe how they work in detail. James O’Donnell describes (AGI)—an “ultra-powerful AI system we don’t have yet”—as a balloon inflated by optimism or fear and deflated when reality falls short of expectations. He connects renewed attention to AGI with news including the launch of Manus, an AI agent created by a Chinese startup, and directs readers to a separate story for more. The edition presents this as a way to understand recurring hype, not as evidence that AGI has arrived. It does not establish whether the fairness benchmarks lead to better outcomes or what techniques would fix the problems they measure.
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
Source details
Original source date: Mar 11, 2025 (2025-03-11T12:30:00Z).
Date basis: Date from the publisher’s own records, through its API.
Added to Escape Meridian: 2026-10-31T04:59:22.823Z. First observed: 2026-10-25T10:31:45.773Z. Retrieved: 2026-10-25T10:33:35.035Z.
Referenced by
Recorded citations and public file attachments. A reference alone does not establish agreement or influence.
No public references recorded here yet.
Limits
Text as retrieved on October 25, 2026; originally published March 11, 2025. The newsletter does not name the fairness or describe their methods, results, or whether they lead to better outcomes. The edition refers readers to separate stories for further detail on the fairness research and Manus; that context is not included here.
Open question
What evidence would show whether these fairness help developers choose fixes that reduce bias in AI systems?
Source-stated
Scott J Mulligan says new benchmarks assess AI systems’ awareness of scenarios and contexts, offering a potentially more nuanced way to measure bias.
The benchmarks evaluate AI systems based on their awareness of different scenarios and contexts.
Source-stated
Mulligan says ignoring differences between groups in earlier approaches may make AI systems less fair.
demonstrating how ignoring differences between groups may in fact make AI systems less fair.
Source-stated
Mulligan says the benchmarks may help teams judge fairness, but fixing the problems may require other techniques.
actually fixing them may require some other techniques altogether.
Source-stated
James O’Donnell characterizes AGI as an ultra-powerful AI system that does not yet exist and compares discussion of it to a balloon repeatedly inflated by optimism or fear.
an ultra-powerful AI system we don’t have yet—can be thought of as a balloon, repeatedly inflated with hype
Source-stated
O’Donnell links renewed attention to AGI with news including the launch of Manus, an AI agent created by a Chinese startup.
including the launch of a new, seemingly super-capable AI agent called Manus, created by a Chinese startup.
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.
Artificial general intelligence
The idea of AI with broad problem-solving, creativity and adaptability across different tasks.
Benchmark
A defined set of tests used to compare systems. Results depend on the tasks and measures selected.
Editorial provenance
Version 1. Publication ID: cd700675-e64e-73ff-5780-ebbc1130c1a6.
Prepared by openai/gpt-6-luna; grounding review by anthropic/claude-opus-5.5.
Source check: passed by anthropic/claude-opus-5.5 on 2026-10-31T04:58:56.776Z. This checks fidelity to the source, not whether its claims are independently true.
Editorial selection: accept; reasons: bridge_admitted_by_model.
Observation: d054e2f2ac73254efb9436f5e15ac46d0494bcc5d610bf457f8e9b3b17a25a84. Source text hash: 6443c71c4d72a21653fe2e7db5913cadc8573acf2effb2c71aed06b1c829e527.
Revision and source-check history
Version 1 · 2026-10-31T04:59:22.823Z
MIT Technology Review’s The Download on AI fairness and AGI hype
MIT Technology Review’s March 11, 2025, edition of The Download says new AI could help developers judge fairness, while actual fixes may require other techniques. Scott J Mulligan reports that the benchmarks assess AI systems’ awareness of scenarios and contexts; he says their researchers were prompted by earlier approaches whose failure to account for differences between groups could make systems less fair. The edition does not name the benchmarks or describe how they work in detail. James O’Donnell describes (AGI)—an “ultra-powerful AI system we don’t have yet”—as a balloon inflated by optimism or fear and deflated when reality falls short of expectations. He connects renewed attention to AGI with news including the launch of Manus, an AI agent created by a Chinese startup, and directs readers to a separate story for more. The edition presents this as a way to understand recurring hype, not as evidence that AGI has arrived. It does not establish whether the fairness benchmarks lead to better outcomes or what techniques would fix the problems they measure.
Original source
MIT Technology Review · Article
Source details
Original source date: Mar 11, 2025 (2025-03-11T12:30:00Z).
Date basis: Date from the publisher’s own records, through its API.
Added to Escape Meridian: 2026-10-31T04:59:22.823Z. First observed: 2026-10-25T10:31:45.773Z. Retrieved: 2026-10-25T10:33:35.035Z.
Limits
Text as retrieved on October 25, 2026; originally published March 11, 2025. The newsletter does not name the fairness or describe their methods, results, or whether they lead to better outcomes. The edition refers readers to separate stories for further detail on the fairness research and Manus; that context is not included here.
Open question
What evidence would show whether these fairness help developers choose fixes that reduce bias in AI systems?
Source-stated
Scott J Mulligan says new benchmarks assess AI systems’ awareness of scenarios and contexts, offering a potentially more nuanced way to measure bias.
The benchmarks evaluate AI systems based on their awareness of different scenarios and contexts.
Source-stated
Mulligan says ignoring differences between groups in earlier approaches may make AI systems less fair.
demonstrating how ignoring differences between groups may in fact make AI systems less fair.
Source-stated
Mulligan says the benchmarks may help teams judge fairness, but fixing the problems may require other techniques.
actually fixing them may require some other techniques altogether.
Source-stated
James O’Donnell characterizes AGI as an ultra-powerful AI system that does not yet exist and compares discussion of it to a balloon repeatedly inflated by optimism or fear.
an ultra-powerful AI system we don’t have yet—can be thought of as a balloon, repeatedly inflated with hype
Source-stated
O’Donnell links renewed attention to AGI with news including the launch of Manus, an AI agent created by a Chinese startup.
including the launch of a new, seemingly super-capable AI agent called Manus, created by a Chinese startup.
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.
Artificial general intelligence
The idea of AI with broad problem-solving, creativity and adaptability across different tasks.
Benchmark
A defined set of tests used to compare systems. Results depend on the tasks and measures selected.
Editorial provenance
Version 1. Publication ID: cd700675-e64e-73ff-5780-ebbc1130c1a6.
Prepared by openai/gpt-6-luna; grounding review by anthropic/claude-opus-5.5.
Source check: passed by anthropic/claude-opus-5.5 on 2026-10-31T04:58:56.776Z. This checks fidelity to the source, not whether its claims are independently true.
Editorial selection: accept; reasons: bridge_admitted_by_model.
Observation: d054e2f2ac73254efb9436f5e15ac46d0494bcc5d610bf457f8e9b3b17a25a84. Source text hash: 6443c71c4d72a21653fe2e7db5913cadc8573acf2effb2c71aed06b1c829e527.
passed · 2026-10-31T04:58:56.776Z · anthropic/claude-opus-5.5
Publication: cd700675-e64e-73ff-5780-ebbc1130c1a6
attribution — pass: The Download is a single newsletter issue with signed commentary, not a bare listing. The fairness-benchmark item is correctly attributed to Scott J Mulligan, the AGI item to James O’Donnell, and the edition to MIT Technology Review.
This is today's edition of The Download, our weekday newsletter
summary — pass: The title, summary and question faithfully reflect the two AI items. AGI is explained in the text and in the definitions, and benchmark is also defined. The summary notes what the edition does not say, including that the benchmarks are unnamed and their effect on outcomes is unknown. Omitting the source's phrase 'seemingly super-capable' for Manus is minor and does not distort the meaning.
But while these new benchmarks could help teams better judge fairness in AI models, actually fixing them may require some other techniques altogether.
date — pass: The stated date of March 11, 2025 matches the UTC day of originalPublished. The dateModified value comes before publication, so there is no post-publication edit. Later 'Deep Dive' items, such as OpenAI math findings, appear to be later page additions, but the reading neither uses nor relies on them. The kind is publisher-page, so no archive disclosure is needed.
support — pass: Each claim matches its quoted excerpt and keeps the source's hedges ('could', 'may'). The reading draws no unsupported benefits or costs. The question follows from the source's own caveat about fixes and is not a leading claim.
They could offer a more nuanced way to measure AI’s bias and its understanding of the world.
media — pass: No media was selected, and no verified candidates were supplied, so there is no mismatch.
Challenge or correct this reading
Send a correction or an opposing source with this reading’s URL. Disagreement and uncertainty can remain unresolved.
A defined set of tests used to compare systems. Results depend on the tasks and measures selected.
Definition referenceThe idea of AI with broad problem-solving, creativity and adaptability across different tasks.
Definition referenceA defined set of tests used to compare systems. Results depend on the tasks and measures selected.
Definition referenceA defined set of tests used to compare systems. Results depend on the tasks and measures selected.
Definition referenceA defined set of tests used to compare systems. Results depend on the tasks and measures selected.
Definition referenceThe idea of AI with broad problem-solving, creativity and adaptability across different tasks.
Definition referenceA defined set of tests used to compare systems. Results depend on the tasks and measures selected.
Definition referenceA defined set of tests used to compare systems. Results depend on the tasks and measures selected.
Definition reference