# MIT Technology Review’s The Download examines OpenAI o1 and 4D printing

[Canonical reading](https://escapemeridian.com/readings/mit-technology-review-s-the-download-examines-openai-o1-and-4d-79c34546)

Published September 17, 2024, MIT Technology Review’s The Download presents two developments: OpenAI’s o1 model and the evolving ambitions of designer and computer scientist Skylar Tibbits. James O’Donnell argues that o1 differs from language-focused models because it is designed for multistep “reasoning” and trained to answer PhD-level questions. He describes it as an early sign that large language models might become useful companions to human researchers, while noting that earlier models have struggled with important problems in fields such as drug discovery, materials science, coding, and physics. The newsletter does not provide test results or other evidence for its comparison of o1 with GPT-4o, so the claimed advance and its research value remain unassessed here.

Anna Gibbs describes Tibbits’s work on 4D printing: materials that can change shape or properties after printing, with time as the “fourth dimension.” She says researchers are exploring possible uses including self-adjusting biomedical devices and soft robotics. Tibbits’s next question is what other capabilities can be built into materials, and whether that can be done without printing. The newsletter does not say what those capabilities might be or whether the potential applications will work in practice.

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 17, 2024 (2024-09-17T12:10:00Z).

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

Added to Escape Meridian: 2026-11-17T04:58:23.756Z. First observed: 2026-10-29T22:58:31.742Z. Retrieved: 2026-10-30T01:32:18.850Z.

- [The Download: OpenAI’s latest model, and 4D printing’s potential](https://www.technologyreview.com/2024/09/17/1104008/the-download-openais-latest-model-and-4d-printings-potential/)

## Limits

The newsletter provides no test results or comparison method supporting its claim that o1 “blows GPT-4o out of the water.” The newsletter describes possible 4D printing applications but gives no evidence that they work in practice. Text as retrieved on October 30, 2026; originally published September 17, 2024. Archived copy.


## Open question

What evidence would show whether o1’s multistep reasoning makes it a useful companion to human researchers on the problems described?


## Source-stated

James O’Donnell says o1 is designed for multistep reasoning and trained to answer PhD-level questions across scientific subjects.

> OpenAI o1 is focused on multistep “reasoning,” the type of process required for advanced mathematics, coding, or other STEM-based questions. The model is also trained to answer PhD-level questions

- [Supporting source](https://www.technologyreview.com/2024/09/17/1104008/the-download-openais-latest-model-and-4d-printings-potential/)

## Source-stated

James O’Donnell presents o1 as an early sign that large language models might become useful companions to human researchers, while saying earlier models have struggled with important problems in several fields.

> OpenAI’s o1 is one of the first signs that LLMs might soon become genuinely helpful companions to human researchers in these fields.

- [Supporting source](https://www.technologyreview.com/2024/09/17/1104008/the-download-openais-latest-model-and-4d-printings-potential/)

## Source-stated

Anna Gibbs says Skylar Tibbits’s 4D printing concept involves materials that change shape or properties after printing; the fourth dimension is time.

> designer and computer scientist Skylar Tibbits started working on 3D-printed materials that could change their shape or properties after being printed—a concept that Tibbits dubbed “4D printing,” where the fourth dimension is time.

- [Supporting source](https://www.technologyreview.com/2024/09/17/1104008/the-download-openais-latest-model-and-4d-printings-potential/)

## Source-stated

Anna Gibbs says researchers are exploring possible applications of 4D printing, including self-adjusting biomedical devices and soft robotics.

> researchers around the world looking into potential applications from self-adjusting biomedical devices to soft robotics.

- [Supporting source](https://www.technologyreview.com/2024/09/17/1104008/the-download-openais-latest-model-and-4d-printings-potential/)

## Source-stated

Anna Gibbs reports that Tibbits is asking what other capabilities can be built into materials and whether this can be done without printing.

> What other capabilities can we build into materials? And can we do that without printing?

- [Supporting source](https://www.technologyreview.com/2024/09/17/1104008/the-download-openais-latest-model-and-4d-printings-potential/)

## 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)

## OpenAI

OpenAI develops AI models and products and publishes research, release announcements and evaluations.

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

## 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: 1857f2c0-d8e9-2eb6-42bd-5d6fd811a24d.

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

Source check: passed by anthropic/claude-opus-5.5 on 2026-11-17T04:54:12.231Z. 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: 78f0a3d7fba35c9e6133eec88e7d8787fb9f7ca98d665cb7bbe994c779361ceb. Source text hash: 5ac58cabc2a50db03ef274955845ab59840c004244dce49b7a41b84631df42a3.


## Revision and source-check history

### Version 1 — 2026-11-17T04:58:23.756Z

Published September 17, 2024, MIT Technology Review’s The Download presents two developments: OpenAI’s o1 model and the evolving ambitions of designer and computer scientist Skylar Tibbits. James O’Donnell argues that o1 differs from language-focused models because it is designed for multistep “reasoning” and trained to answer PhD-level questions. He describes it as an early sign that large language models might become useful companions to human researchers, while noting that earlier models have struggled with important problems in fields such as drug discovery, materials science, coding, and physics. The newsletter does not provide test results or other evidence for its comparison of o1 with GPT-4o, so the claimed advance and its research value remain unassessed here.

Anna Gibbs describes Tibbits’s work on 4D printing: materials that can change shape or properties after printing, with time as the “fourth dimension.” She says researchers are exploring possible uses including self-adjusting biomedical devices and soft robotics. Tibbits’s next question is what other capabilities can be built into materials, and whether that can be done without printing. The newsletter does not say what those capabilities might be or whether the potential applications will work in practice.

## Original source

MIT Technology Review · Article

Original source date: Sep 17, 2024 (2024-09-17T12:10:00Z).

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

Added to Escape Meridian: 2026-11-17T04:58:23.756Z. First observed: 2026-10-29T22:58:31.742Z. Retrieved: 2026-10-30T01:32:18.850Z.

- [The Download: OpenAI’s latest model, and 4D printing’s potential](https://www.technologyreview.com/2024/09/17/1104008/the-download-openais-latest-model-and-4d-printings-potential/)

## Limits

The newsletter provides no test results or comparison method supporting its claim that o1 “blows GPT-4o out of the water.” The newsletter describes possible 4D printing applications but gives no evidence that they work in practice. Text as retrieved on October 30, 2026; originally published September 17, 2024. Archived copy.


## Open question

What evidence would show whether o1’s multistep reasoning makes it a useful companion to human researchers on the problems described?


## Source-stated

James O’Donnell says o1 is designed for multistep reasoning and trained to answer PhD-level questions across scientific subjects.

> OpenAI o1 is focused on multistep “reasoning,” the type of process required for advanced mathematics, coding, or other STEM-based questions. The model is also trained to answer PhD-level questions

- [Supporting source](https://www.technologyreview.com/2024/09/17/1104008/the-download-openais-latest-model-and-4d-printings-potential/)

## Source-stated

James O’Donnell presents o1 as an early sign that large language models might become useful companions to human researchers, while saying earlier models have struggled with important problems in several fields.

> OpenAI’s o1 is one of the first signs that LLMs might soon become genuinely helpful companions to human researchers in these fields.

- [Supporting source](https://www.technologyreview.com/2024/09/17/1104008/the-download-openais-latest-model-and-4d-printings-potential/)

## Source-stated

Anna Gibbs says Skylar Tibbits’s 4D printing concept involves materials that change shape or properties after printing; the fourth dimension is time.

> designer and computer scientist Skylar Tibbits started working on 3D-printed materials that could change their shape or properties after being printed—a concept that Tibbits dubbed “4D printing,” where the fourth dimension is time.

- [Supporting source](https://www.technologyreview.com/2024/09/17/1104008/the-download-openais-latest-model-and-4d-printings-potential/)

## Source-stated

Anna Gibbs says researchers are exploring possible applications of 4D printing, including self-adjusting biomedical devices and soft robotics.

> researchers around the world looking into potential applications from self-adjusting biomedical devices to soft robotics.

- [Supporting source](https://www.technologyreview.com/2024/09/17/1104008/the-download-openais-latest-model-and-4d-printings-potential/)

## Source-stated

Anna Gibbs reports that Tibbits is asking what other capabilities can be built into materials and whether this can be done without printing.

> What other capabilities can we build into materials? And can we do that without printing?

- [Supporting source](https://www.technologyreview.com/2024/09/17/1104008/the-download-openais-latest-model-and-4d-printings-potential/)

## 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)

## OpenAI

OpenAI develops AI models and products and publishes research, release announcements and evaluations.

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

## 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: 1857f2c0-d8e9-2eb6-42bd-5d6fd811a24d.

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

Source check: passed by anthropic/claude-opus-5.5 on 2026-11-17T04:54:12.231Z. 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: 78f0a3d7fba35c9e6133eec88e7d8787fb9f7ca98d665cb7bbe994c779361ceb. Source text hash: 5ac58cabc2a50db03ef274955845ab59840c004244dce49b7a41b84631df42a3.


### passed — 2026-11-17T04:54:12.231Z — anthropic/claude-opus-5.5

Publication: 1857f2c0-d8e9-2eb6-42bd-5d6fd811a24d

attribution — pass: This is one newsletter issue that comments on its stories in its own words, not a bare listing. The o1 item is correctly attributed to James O’Donnell and the 4D printing item to Anna Gibbs, matching the bylines. The publisher, MIT Technology Review’s The Download, is named.

> —James O’Donnell

summary — pass: The reading leads with the two featured developments and names the actors. It explains multistep reasoning and 4D printing in plain terms, and the page definitions cover large language model. It keeps O’Donnell’s hedged framing (“might soon”) and notes that the newsletter gives no evidence for the comparison with GPT-4o. The question is open, not leading. Leaving out the unrelated must-reads list is reasonable.

> OpenAI’s o1 is one of the first signs that LLMs might soon become genuinely helpful companions to human researchers in these fields.

date — pass: September 17, 2024 matches the UTC day of originalPublished. The limitations disclose the archived copy and the later retrieval. The reading presents nothing as happening after publication. The o1 “companion” idea is kept as an attributed possibility, not an outcome. The archive capture on the same day as publication raises no concern about later edits.

> Last week OpenAI released a new model called o1

support — pass: Each claim matches the quoted text. Saying earlier models “struggled” fairly paraphrases “failed to demonstrate the types of skills required.” The applications are framed as possible uses being explored, not demonstrated benefits. The limitations accurately note that the newsletter gives no test evidence. The reading adds no unsupported benefits or costs.

> such LLMs have failed to demonstrate the types of skills required to solve important problems in fields like drug discovery, materials science, coding, or physics

media — pass: No media is selected and no verified candidates were supplied, so nothing was invented or misidentified.

## 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/79c3454642a4e075eb045b96028bada3440300f12164fd656f46c8e933234ffe/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-s-the-download-examines-openai-o1-and-4d-79c34546.json)
