Scores by kind of claim
Each kind of claim is scored on its own, out of 100: the share that held up, with the range it likely falls in. Fewer resolved claims give a wider range.
-
Factual claims
Statements about the present or the past, checked against the record: did they hold up?
80 74 86 -
Predictions
Statements about events outside the person's control, checked once they fell due: did they come true?
Insufficient record: 4 predictions resolved, 5 needed for a score.
-
Promises
Commitments to the person's or organisation's own future action, with something to deliver: was it delivered?
85 74 94 -
Guidance
A company's own figures for its results over a period, checked against what it then reported.
Insufficient record: 0 guidance claims resolved, 5 needed for a score.
All types combined: 81 (likely 76 to 86, high confidence), from 86 claims: 70 factual claims, 16 promises. It mixes kinds of claim in different proportions for different people, so compare people kind by kind.
Mark Zuckerberg's claims
40 claims
“One that people are going to find pretty wild is that it now generates high quality images so quickly that it actually generates it as you're typing and updates it in real time.”
Mark Zuckerberg, Meta CEO 18 April 2024 Mark Zuckerberg - Llama 3, Open Sourcing $10b Models, & Caesar Augustus (Dwarkesh Podcast)
Meta AI now generates high-quality images fast enough to generate and update them in real time as the user types.
Mostly true True in substance, wrong in a detail.
Said for an organisation they lead or work for; it counts on both records.
What would count as true or false
True: as of 18 April 2024, Meta AI had a live feature (for example 'Imagine' with real-time generation) that generated images and updated them as the user typed, as shown by Meta announcements, demos or hands-on reviews. Mostly true: the feature existed but was in limited rollout (for example only some regions or only in beta) at launch. Mostly false: it had only been announced, or the images updated only after the user finished typing. False: there was no such capability.
Evidence
- Meet Your New Assistant: Meta AI, Built With Llama 3 Meta (about.fb.com newsroom) · 18 April 2024 Own record
“We’re making image generation faster, so you can create images from text in real-time using Meta AI’s Imagine feature. We’re starting to roll this out today in beta on WhatsApp and the Meta AI web experience in the US.”
- Meet Your New Assistant: Meta AI, Built With Llama 3 Meta (about.fb.com newsroom) · 18 April 2024 Own record
“You’ll see an image appear as you start typing — and it’ll change with every few letters typed, so you can watch as Meta AI brings your vision to life.”
- Meta’s Imagine AI image generator just got a big GIF upgrade — and I’m obsessed Tom's Guide (via inkl) · 19 April 2024
“Imagine now comes with a live generation capability, where as you type the image generates in front of you, changing with each new word — up to a limit.”
- Meta AI's Image Generation Is Unlike Anything ChatGPT Can Do - BGR BGR · 19 April 2024
“However, the Imagine feature is restricted to the US market and only available on WhatsApp and the web (over at Meta.ai). Image generation is also in beta at the moment, so you might encounter issues along the way.”
Reasoning
Meta's launch announcement says real-time image generation in Imagine began rolling out on 18 April 2024, with images appearing as the user types and changing every few letters (E1, E2). A Tom's Guide hands-on review the next day confirms that images were generated live as the user typed, changing with each new word up to a limit (E3). However, the feature was in beta and limited to the US on WhatsApp and the web (E1, E4). The criteria class that kind of limited rollout as mostly true.
Cite this claim
Mark Zuckerberg, “Meta AI now generates high-quality images fast enough to generate and update them in real time as the user types.” (said 18 April 2024). Verdict: mostly true. Reliability API, claim c_62ee49aec520, method 0.1.0+3f84d14d. https://epistacklabs.com/claims/c_62ee49aec520
Claims without a verdict
- Unresolvable 144
- Not yet due 17
- Own records only 41
- Not researched 223
- Not checkable 1,060
Is a verdict here wrong, or do you speak for Mark Zuckerberg? Reply on the claim's own page, with the form at its foot (how replies work). Verdicts are made automatically; how verdicts and scores are made.