“The two megacompanies both demoed prototypes, neither fully ready for public use, built around apparently comparable technology, facing apparently similar bugs, within a day of each other. Yet one demo was presented as a revolution, the other as a disaster.”
Our reading:
Google and Microsoft both demoed prototypes, neither fully ready for public use, built around apparently comparable technology and facing apparently similar bugs, within a day of each other, yet Microsoft's demo was presented as a revolution and Google's as a disaster.
Assessment: Supported
What happened
Was it true for the period described?
Outcome: Supported
The evidence matches what they said.
Judged on a plain reading. They named no measure, so we used the reading a typical informed reader would take. It is spelled out under “How we judged it”.
Evidence up to . Checked .
Why
Microsoft demonstrated Bing on February 7, 2023, and Google held its Paris presentation on February 8 [e1, e3]. Microsoft described a limited preview using an OpenAI language model; Google's February 6 announcement described Bard as an experimental LaMDA service opening to trusted testers [e1, e2]. These records support the comparison of language-model tools with restricted access and unresolved reliability problems [e1, e2, e4]. AP's February 8 report documented Bard's false telescope answer and Bing's invented result for a Super Bowl that had yet to occur [e4]. Bard's error appeared in its earlier advertisement [e3]. Ars Technica sharply criticized Google's Paris presentation, while a February 9 roundup reported praise for Bing as a major change in accessing information [e3, e6]. These examples support the reception contrast within February 6–9, although AP also criticized Bing and a favorable attendee review credited its integration [e3, e4, e5, e6].
Caveats
- The coverage examples do not establish how common either reaction was or show unanimous reception [e3–e6].
- Restricted access and documented errors do not provide an objective test of public readiness or establish equal accuracy [e1, e2, e4].
- The event records establish adjacent calendar dates, without establishing a strict 24-hour interval [e1, e3].
- The February 9 roundup supports reception within the target period; it does not establish that this roundup was available before Marcus published [e6].
Evidence (6 items)
Supports the claim: Microsoft,
a desktop limited preview.
Why it matters: The dated transcript records Microsoft's demonstration, restricted preview and use of a next-generation OpenAI language model. These support the timing, broad technology comparison and unfinished rollout described in the claim.
- Publisher
- Microsoft
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- About events on
- Published
Supports the claim: Google,
an experimental conversational AI service
Why it matters: Google describes a language-model-based experimental service and says access is opening to trusted testers before wider public availability. Its testing plans support the ordinary reading of incomplete public readiness.
- Publisher
- Kind of source
- Company disclosure: the company describing itself, not an independent finding
- About events on
- Published
Supports the claim: Ars Technica,
an AI fizzle.
Why it matters: This contemporary report supplies the Paris event date and strongly negative reception. It distinguishes Bard's earlier advertisement error from the presentation and reports that Bard remained restricted to testing.
- Publisher
- Ars Technica
- Kind of source
- News report
- About events on
- Published
Supports the claim: Associated Press, syndicated by Tech Xplore,
spitting out falsehoods.
Why it matters: AP documents Bard's telescope error and its own test in which Bing invented a completed Super Bowl before the game occurred. This supports similar factual-error problems while showing that contemporary coverage also criticized Bing.
- Publisher
- Associated Press, syndicated by Tech Xplore
- Kind of source
- News report
- About events on
- Published
Supports the claim: Search Engine Roundtable,
far more impressive
Why it matters: An attendee explicitly favors Microsoft's demonstration over Google's initial Bard showing. His account identifies integration and user experience as reasons, so the reception difference need not imply that the underlying models differed greatly.
- Publisher
- Search Engine Roundtable
- Kind of source
- News report
- About events on
- Published
Supports the claim: Newser,
Roose isn't alone in his praise.
Why it matters: The article documents reviewers presenting Bing as a major change in accessing information, supporting the ordinary meaning of revolutionary reception. This is secondary evidence because the New York Times page could not be read directly.
- Where
- Second paragraph; surrounding summaries of New York Times, Wall Street Journal and Inverse reviews
- Publisher
- Newser
- Kind of source
- Secondary summary
- About events on
- Published
How we judged it
- Kind of claim
- Observation
- Deadline
- None given. We don’t add one.
Conditions
- apparently comparable technology (their words)
- apparently similar bugs (their words)
- neither fully ready for public use (their words)
Where we had to interpret
- The target is the launch-week demonstrations and reception discussed in the February 9 post, covering February 6–9, 2023.
- Within a day means demonstrations on adjacent calendar dates: Microsoft's February 7 event and Google's February 8 Paris event. Google's February 6 Bard announcement is distinguished from its Paris presentation.
- Prototypes and not fully ready mean experimental or preview systems with restricted access and unresolved reliability problems. This does not mean nobody could use them.
- Apparently comparable technology means language-model-based conversational information tools, without requiring identical models, integrations, capabilities, or accuracy.
- Apparently similar bugs means factual errors presented as answers, without requiring equal error rates.
- The reception contrast means prominent contemporary examples of transformative praise for Bing and sharply negative treatment of Google. It does not require unanimous coverage or those exact labels.
Right if: Records establish adjacent-day demonstrations, experimental or preview status, broadly comparable language-model technology and factual-error problems, together with prominent contemporary examples of the stated reception contrast.
Wrong if: Reliable records materially refute the timing, developmental status, broad technological or error similarity, or claimed reception contrast. Differences in implementation alone do not refute apparent comparability.
We wrote this test after the outcome was known.
We found 9 claims in this post. See them all.
Full record
Source and dates
Marcus on AI, Oops! How Google bombed, while doing pretty much exactly the same thing as Microsoft did, with similar results, paragraph 12.
- Said
- We read the source
- live page, full text, public
- Evidence up to
- Checked
We read this wording on 2026-10-29. The post may have been edited after 2023-02-09.
Where it was said (1)
Said once in the posts we read.
: Marcus on AI, Oops! How Google bombed, while doing pretty much exactly the same thing as Microsoft did, with similar results
The two megacompanies both demoed prototypes, neither fully ready for public use, built around apparently comparable technology, facing apparently similar bugs, within a day of each other. Yet one demo was presented as a revolution, the other as a disaster.
- Where
- paragraph 12
- Published
- 2023-02-09
- We read it
- , live page
- Access
- full text, public
History
Versions
- Current: Supported. Evidence up to 2026-10-28.
What happened to this record
PublishedBy us
Published after both AI reviews and the final check passed.
Held backBy us
The final check failed: the write-up said more than the evidence showed.