Welp, Gallup is no longer a source of viable anything, let along survey data:

[mildpeach@mstdn.social]

@octothorpe @JonChevreau This is bad. No matter what we think of AI slop, actual human driven chaos is worse. That is why there is not modeling it…

[oliver_schafeld@mastodon.online]

"Due to cost reasons" has gotta be a forward-looking statement translating as "we're afraid to be sued or regulated out of business by you-know-who" if we publish polls about you-know-who.

[mast0d0nphan@beige.party]

@octothorpe @gbargoud Isn't that what one of the pilots on the recent moon mission said?

[d_rift@beige.party]

@octothorpe I expected a venerable polling company at least, unlike the rest of the modern world, to understand the fundamental implications of basic statistical sampling methods on data validity. Alas.

[stekopf@mstdn.social]

@octothorpe

»Gallup says it is exploring "digital twin" technology that replaces human respondents with AI/LLM-generated answers«

What #Gallup says instead:
»Simulated responses will not be used to produce Gallup’s published population estimates, and they will not replace direct measurement of people in our tracking research. […] We will always be transparent about our use of simulated responses and will never present them as human responses.«
From the linked text.

The critical point is…

1/2

[stekopf@mstdn.social]

…that they plan to filter research:
»Simulated approaches may also expand what is practically researchable, enabling organizations to explore attitudinal questions at a scale or speed that would otherwise be prohibitive, to model responses among hard-to-reach populations, or to test how different groups might react to novel stimuli before investing in full-scale data collection. In short, agent banks may help researchers and practitioners anticipate likely outcomes and support decision-making.«

[ariaflame@masto.ai]

@octothorpe ‍️

[octothorpe@mastodon.online]

@noodlemaz thanks!

[octothorpe@mastodon.online]

@d_rift they have a lot of name recognition but often their methodologies are questionable, allowing the sponsor of the study to kinda numberwang the results they want to promote.

[d_rift@beige.party]

@octothorpe still; thanks for the heads-up on them moving from numberwang to "no number, all wang" territory.