# MIT Technology Review on AI moving beyond chat—and the search for a killer app

[Canonical reading](https://escapemeridian.com/readings/mit-technology-review-on-ai-moving-beyond-chat-and-the-search-802cc609)

Published October 8, 2024, MIT Technology Review describes AI tools that work with images, audio, and text, and interfaces that let people edit or search through material in new ways. The article points to Google’s NotebookLM Audio Overview, Meta’s Movie Gen, OpenAI’s Canvas, and Google’s video-and-voice search as examples. The author says NotebookLM’s unexpected appeal brought “wonder and delight,” while its success as a side feature suggests AI developers do not know what people will want.

The piece argues that improved generative AI content and pressure on AI companies to make money have produced a wave of applications, but not a “killer app.” Its central tension is between the striking range of new interactions and uncertainty about which tools will matter to users. The author characterizes the current approach as companies throwing different tools at people and seeing what sticks. The article gives examples and impressions, but does not establish how widely the features are used or whether they produce lasting value. It leaves open whether any of these applications will become a widely useful product, rather than another experiment in a crowded field.

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: Oct 8, 2024 (2024-10-08T09:04:52Z).

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

Added to Escape Meridian: 2026-11-17T00:43:23.322Z. First observed: 2026-10-30T09:58:32.231Z. Retrieved: 2026-10-30T10:00:08.166Z.

- [Forget chat. AI that can hear, see, and click is already here.](https://www.technologyreview.com/2024/10/08/1105214/forget-chat-ai-that-can-hear-see-and-click-is-already-here/)

## Limits

Text as retrieved on October 30, 2026; originally published October 8, 2024. The article gives examples and the author’s impressions, but does not provide usage measurements or evidence of lasting value for the applications discussed.


## Open question

What evidence would show that one of these multimodal applications has become a widely useful product rather than a feature people try briefly?


## Source-stated

MIT Technology Review describes multimodal models and applications that use different kinds of source material and offer more interactive interfaces.

> The latest hot AI toys take advantage of multimodal models, which can handle several things at the same time, such as images, audio, and text.

- [Supporting source](https://www.technologyreview.com/2024/10/08/1105214/forget-chat-ai-that-can-hear-see-and-click-is-already-here/)

## Source-stated

The author says NotebookLM’s Audio Overview brought “wonder and delight” and that its popularity as a side feature suggests AI developers do not know what users will want.

> But the fact that NotebookLM’s Audio Overviews became a hit despite being a side feature hidden inside a bigger product just goes to show that AI developers don’t really know what they are doing.

- [Supporting source](https://www.technologyreview.com/2024/10/08/1105214/forget-chat-ai-that-can-hear-see-and-click-is-already-here/)

## Source-stated

The author argues that despite rapid improvement in generative AI content, there has not yet been a killer app, and companies are testing tools under pressure to make money.

> But we’ve yet to see a killer app, and these new multimodal applications are a result of the immense pressure AI companies are under to make money and deliver.

- [Supporting source](https://www.technologyreview.com/2024/10/08/1105214/forget-chat-ai-that-can-hear-see-and-click-is-already-here/)

## Source-stated

The author characterizes AI companies’ current approach as trying different tools with users to see which ones catch on.

> Tech companies are throwing different AI tools at people and seeing what sticks.

- [Supporting source](https://www.technologyreview.com/2024/10/08/1105214/forget-chat-ai-that-can-hear-see-and-click-is-already-here/)

## 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.


## Multimodal model

A model that works with more than one kind of input or output, such as text, images and audio.

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

## Editorial provenance

Version 1. Publication ID: 6bbabf43-69a0-060b-0f53-69929b6b74f8.

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

Source check: uncertain by anthropic/claude-opus-5.5 on 2026-11-17T00:38:43.692Z. 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: 4e80736fe139659264dfc30e074fa63e1456d862f6bd14e36e4f8b532fa77215. Source text hash: a15c0d70e3d87bc4f5b30a73d44752f576ddf9891bf33f3ef2da49444c9288b7.


## Revision and source-check history

### Version 1 — 2026-11-17T00:43:23.322Z

Published October 8, 2024, MIT Technology Review describes AI tools that work with images, audio, and text, and interfaces that let people edit or search through material in new ways. The article points to Google’s NotebookLM Audio Overview, Meta’s Movie Gen, OpenAI’s Canvas, and Google’s video-and-voice search as examples. The author says NotebookLM’s unexpected appeal brought “wonder and delight,” while its success as a side feature suggests AI developers do not know what people will want.

The piece argues that improved generative AI content and pressure on AI companies to make money have produced a wave of applications, but not a “killer app.” Its central tension is between the striking range of new interactions and uncertainty about which tools will matter to users. The author characterizes the current approach as companies throwing different tools at people and seeing what sticks. The article gives examples and impressions, but does not establish how widely the features are used or whether they produce lasting value. It leaves open whether any of these applications will become a widely useful product, rather than another experiment in a crowded field.

## Original source

MIT Technology Review · Article

Original source date: Oct 8, 2024 (2024-10-08T09:04:52Z).

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

Added to Escape Meridian: 2026-11-17T00:43:23.322Z. First observed: 2026-10-30T09:58:32.231Z. Retrieved: 2026-10-30T10:00:08.166Z.

- [Forget chat. AI that can hear, see, and click is already here.](https://www.technologyreview.com/2024/10/08/1105214/forget-chat-ai-that-can-hear-see-and-click-is-already-here/)

## Limits

Text as retrieved on October 30, 2026; originally published October 8, 2024. The article gives examples and the author’s impressions, but does not provide usage measurements or evidence of lasting value for the applications discussed.


## Open question

What evidence would show that one of these multimodal applications has become a widely useful product rather than a feature people try briefly?


## Source-stated

MIT Technology Review describes multimodal models and applications that use different kinds of source material and offer more interactive interfaces.

> The latest hot AI toys take advantage of multimodal models, which can handle several things at the same time, such as images, audio, and text.

- [Supporting source](https://www.technologyreview.com/2024/10/08/1105214/forget-chat-ai-that-can-hear-see-and-click-is-already-here/)

## Source-stated

The author says NotebookLM’s Audio Overview brought “wonder and delight” and that its popularity as a side feature suggests AI developers do not know what users will want.

> But the fact that NotebookLM’s Audio Overviews became a hit despite being a side feature hidden inside a bigger product just goes to show that AI developers don’t really know what they are doing.

- [Supporting source](https://www.technologyreview.com/2024/10/08/1105214/forget-chat-ai-that-can-hear-see-and-click-is-already-here/)

## Source-stated

The author argues that despite rapid improvement in generative AI content, there has not yet been a killer app, and companies are testing tools under pressure to make money.

> But we’ve yet to see a killer app, and these new multimodal applications are a result of the immense pressure AI companies are under to make money and deliver.

- [Supporting source](https://www.technologyreview.com/2024/10/08/1105214/forget-chat-ai-that-can-hear-see-and-click-is-already-here/)

## Source-stated

The author characterizes AI companies’ current approach as trying different tools with users to see which ones catch on.

> Tech companies are throwing different AI tools at people and seeing what sticks.

- [Supporting source](https://www.technologyreview.com/2024/10/08/1105214/forget-chat-ai-that-can-hear-see-and-click-is-already-here/)

## 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.


## Multimodal model

A model that works with more than one kind of input or output, such as text, images and audio.

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

## Editorial provenance

Version 1. Publication ID: 6bbabf43-69a0-060b-0f53-69929b6b74f8.

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

Source check: uncertain by anthropic/claude-opus-5.5 on 2026-11-17T00:38:43.692Z. 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: 4e80736fe139659264dfc30e074fa63e1456d862f6bd14e36e4f8b532fa77215. Source text hash: a15c0d70e3d87bc4f5b30a73d44752f576ddf9891bf33f3ef2da49444c9288b7.


### uncertain — 2026-11-17T00:38:43.692Z — anthropic/claude-opus-5.5

Publication: 6bbabf43-69a0-060b-0f53-69929b6b74f8

attribution — pass: This is a single newsletter column from The Algorithm, written in the author's own words, not a listing. The reading attributes it to MIT Technology Review and 'the author'. The author is not named in the captured text, and the reading does not invent a name.

> This story originally appeared in The Algorithm, our weekly newsletter on AI.

summary — uncertain: The reading softens the author's pointed line into 'AI developers do not know what people will want' (summary) and 'what users will want' (claim 2). The author says developers 'don't really know what they are doing.' Smallest repair: quote that phrase directly. The examples are also only named. Audio Overview (AI-generated podcasts from your material), Movie Gen (text-to-video) and Canvas (selecting text or code to edit instead of chatting) are not described, so a general reader may not see what changed. One plain clause each would fix this.

> just goes to show that AI developers don’t really know what they are doing

date — pass: October 8, 2024 matches the UTC day of originalPublished (2024-10-08T09:04:52Z). The page is a publisher page, so no archive disclosure is needed. The Deep Dive sidebar refers to later events, such as 'Ten years after AlphaGo'. These are possible later additions, and the reading neither uses nor relies on them. No forecasts are presented as outcomes.

support — pass: The claims about multimodal models, the missing killer app, pressure to make money, and 'seeing what sticks' are directly stated in the source. The stated limitation is accurate: the piece gives no usage data. The question follows from the source's own uncertainty and does not assume an outcome. The softened paraphrase in claim 2 is noted under summary.

> Tech companies are throwing different AI tools at people and seeing what sticks.

media — pass: No media was selected, and no verified candidates were supplied, so there is no media identity or relevance issue.

## 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/802cc609d05561a32fdce3eff0d8069fdbfc0ffff9c196a030928befb20140ad/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-on-ai-moving-beyond-chat-and-the-search-802cc609.json)
