Whenever I see a claim that "x% of adults do y", my brain goes:
- "x% of what? what is the denominator?". Without that number, the claim is meaningless.
- surely it cannot be the entire population, so it has to be a survey.
- how many people participated in the survey? what was the distribution?
Here is that info for this study. I found this in the PDF version of the study report [0] referred to at the end of the Northwestern page [1].
> Methodology
The Harris Poll conducted a total of 4,375 online interviews among the general U.S. adult (18+) population between January 5th and January 21st, 2026. Included in this overall total is a sample of 816 High-Net-Worth individuals (those with total household investable assets, excluding pensions, retirement plans and property, greater than $1,000,000).
I love mental rule of thumbs / heuristics that we can install into our brain to avoid getting caught up in cognitive biases or other mistakes.
An easy one that I would expect most HN readers probably do already is: When shopping, always round up to the nearest dollar before even mentally storing it or operating on it. The usual cognitive bias is that many people end up storing the listed price of $4.96 in a lossy manner, as $4.xx, and end up thinking of it as $4 when in reality they could be skipping straight to keeping it as $5 in their head.
it blows my mind when anyone ever reads $x.99 as $x out loud. i often cant keep myself from muttering "$x+1" and its taken decades of practice to keep myself from looking at them like they have 2 heads
figured it out since i learned to read so it seems so childish to hear adults make it
turns out ppl like me are the weird ones and most people just truncate at the period
You are reading gp’s comment correctly as it was written, but it omits the next line in the source study:
> Data for the general U.S. population (including the High Net Worth oversample) were weighted to
Census targets for education, age, gender, race/ethnicity, region and household income.
They oversampled in major markets where they work and in high-net-worth populations (who they service), but their claims are for the overall US adult population.
Oversampling like this is pretty routine in survey research. It improves the precision of any subgroup analyses you might want to do, and, to a first approximation, it doesn’t tend to bias the weighted overall-population claims in one direction or another.
I think about it like Google Earth or something. I happen to have much-higher-res imagery of London than of the Cotswolds. That doesn’t mean my view, when zoomed out to “the whole United Kingdom,” is necessarily misleading. It does mean I can additionally make more detailed claims about Piccadilly Circus than about the sheep fields or whatever.
Typically surveys are adjusted for sampling biases before reporting. That appears to be the case here. So there is usually some attempt to account for the biases in the sampled population.
The impulse to ask "what population was sampled?" is good but its not always a straight line from there to "these results directly reflect that sampling bias."
In fact, from the page you posted: "Data for the general U.S. population (including the High Net Worth oversample) were weighted to Census targets for education, age, gender, race/ethnicity, region and household income. A full methodology is available."
I would presume that the headline number attempts to account for sampling bias.
I agree with you - there’s usually some adjustment for sampling bias, and this study says it is matching Census targets. But I had to go three levels down to see that info before I derive any meaning from the number.
My concern is that headlines like “x% of adults do y” get repeated without anyone (sometimes even journalists writing the article) seeing the methodology or nuance behind them. Context matters.
Europe in general maybe outside of some CEE, DACH, Nordics/Benelux is pretty fucked for the youth. Without help from parents or inherited housing you're fucked. You get American CoL with African wages. Joking of course, but you get the point.
Another reason the denominator is important is that "parents" are (hopefully) completely a subset of "adults". And not all adults have living parents. So who is counted:
- All adults
- Adults with living parents
- Adults within an age range
- Adults without children
or what? This really belongs in the title. Without it, as you say, the statistic is meaningless.
My follow-up question is "What percentage of the 'totally independent' group, are parents with adult children?"
i.e. there's almost an interesting statistic here about what percentage of adults have "no option" in relying on their parents for financial help, and what percentage of the remainder still do rely on their parents.
But once I had this question, I realized that I needed to be suspicious of the % itself.
Related - any time a study is based on gathering the opinions of a random subset of the population, I also instantly dismiss it. The average person is a moron. I don't care about random people's subjective opinions, I only care about objective data. People polled in the 16th century would have said the sun orbits the earth; that doesn't make it true.
Apparently it is Language Model, as mentioned in the announcement of NotebookLM in 2023 [0].
> Today we’re beginning to roll out Project Tailwind with its new name: NotebookLM, an experimental offering from Google Labs. It’s our endeavor to reimagine what notetaking software might look like if you designed it from scratch knowing that you would have a powerful language model at its core: hence the LM.
It's funny how similar that article's intro is to today's announcement.
In fact, it would almost certainly encourage them to keep you engaged even longer in an attempt to make up some of the money lost with the end of targeted advertising.
Wouldn’t it lead to the opposite? You spending time on the platform earns them money because they’re gathering data for targeted ads, and showing you those ads. Non-targeted ads barely pay anything.
Addictive content feeds are expensive with the live HD video playing everywhere and the constant tweaks needed from teams working to further refine targeting based on behavior.
For typical social media sites, engagement will pretty much always be proportional to both revenue and cost per user. Either revenue > cost, and the site is incentivized to increase engagement, or revenue < cost and the site dies. There is no middle ground where a site gets a healthy revenue that's greater than its costs, but increasing engagement won't increase revenue. The exceptions are niche sites that do things like fixed subscriptions, or cost money to create content but not to consume it (but even in these cases, increasing engagement probably still increases the chance users start/continue to be paid customers).
You assume that subscriptions wouldn’t become more common without targeted ads. It’s certainly possible.
(I’m not sure if it’s even possible to ban targeted ads, haven’t thought much about it. Perhaps there’s a regulation of commerce angle. I do think that businesses could be forced to provide more clarity about the exchange that’s taking place, the value of the data, how the data is used, and so on.)
I agree it’s very possible subscriptions would become more common if targeted ads were restricted. Personally I’d rather pay money for a good product/service than 'being the product' for a bad product that monetizes my data.
Unlike most articles that butcher an analogy so badly that you wish they could have just described the concept plainly, this one uses the analogy really well. It carries it from start to finish without overstretching it.
This line captures the essence of the article and is going to stick with me forever:
> SaaS is the bread, not the bread machine.
And yes, SaaS companies that understand that they sell convenience and accountability will be the ones that survive this AI rush. New ones could emerge too.
Such a great read! I kept on nodding and chuckling the whole time reading it. I can see myself as the founder, especially 'spending time in the oven forums' lol.
I went to the /blog route to see other posts by the author, but alas, there is only this one! And that's a gem.
I THINK I fixed it... should be up in 10 minutes. The reason: the try.piecesof.me flow creates a profile in a sort of exists-but-also-doesnt-exists state. Made a change so that the doc is viewable in that middle state. But anyone who claimed it, before and now, would've been able to see the doc.
Thanks for trying it out though and I hope to see you on the platform.
Thanks for the note.
While building the context bundle, there is a tool that detect gaps in the timeline. So when you ask 'kitchen renovation', and there are emails from 2019 and 2023, the tool flags these gaps and the LLM is made aware of it. So Memento asks which of these clusters you want to track as a project.
Later while generating the wiki, LLM handles conflicts and facts that evolved over time because it has the timestamp of the messages. So it shows that in the narrative.
Wwat we haven't implemented yet is to update the wiki when new emails arrive. That is the next one in our backlog.
- "x% of what? what is the denominator?". Without that number, the claim is meaningless. - surely it cannot be the entire population, so it has to be a survey. - how many people participated in the survey? what was the distribution?
Here is that info for this study. I found this in the PDF version of the study report [0] referred to at the end of the Northwestern page [1].
> Methodology The Harris Poll conducted a total of 4,375 online interviews among the general U.S. adult (18+) population between January 5th and January 21st, 2026. Included in this overall total is a sample of 816 High-Net-Worth individuals (those with total household investable assets, excluding pensions, retirement plans and property, greater than $1,000,000).
[0] https://filecache.mediaroom.com/mr5mr_nwmutual/179168/2026%2...
[1] https://news.northwesternmutual.com/planning-and-progress-st...