Feeds reward whoever sounds most certain, which tells you little about who is right. What gets lost in the short version, and how to find people who know.

By Will Nash
18 November 2026
If you have tried to learn a subject from what your feeds serve up, you will have noticed that the most confident voice tends to get the most room. That is a result of how the feed is built, and it tells you little about who is right.
Ranking systems reward what people respond to quickly, and certainty does well in a few seconds while a careful caveat does not. To find the people who actually know a subject, you have to look for things a feed has no good way of rewarding, and give them more time than a clip allows.
What a feed is built to reward
Most social platforms decide what to show you by predicting what you will engage with: whether you will watch to the end, reply, share or keep scrolling. That prediction is built from behaviour measured in seconds.
A fast, confident claim does well by that measure, and so does a provocative one. An answer that opens with "it depends" usually does not, even when "it depends" is the accurate place to start.
None of this requires anyone to be dishonest. The people making confident clips are often sincere, and the systems ranking them are doing what they were designed to do. Accuracy is very hard to measure in the moment, though, so it carries far less weight in the ranking than reaction does, and confidence and correctness end up sorted together.
What gets lost in the short version
Specialists spend most of their working lives in the parts of a subject that do not fit in a clip, such as the conditions under which a finding holds and the cases that break the rule.
Take one of the questions we are asking at Aloudable: what the replication crisis means for a drug already on the market. A thirty-second answer has room for one position, roughly "the evidence is broken" or "there is nothing to worry about". A useful answer has to deal with what the original approval rested on and what has been checked since, and that takes longer to say.
When someone who knows a field well says "we do not know yet", they are telling you something useful about the state of the evidence, and in a feed the same sentence tends to read as weakness. The person who sounds sure gets the next share.
How to find the people who actually know
A later post in this series goes further into judging expertise, but a few habits help straight away.
Look for time spent on one subject. Someone who has worked on one narrow question for years will usually tell you more about it than a commentator who covers whatever is trending this week.
Notice who volunteers the limits of their own claim. Being told where a finding stops applying is a good sign, because you can only know that if you know the finding well.
Prefer specifics you could check, such as a named study or a described mechanism, over a general statement of how things are.
Ask who chose the speaker. A feed chose the clip because people reacted to it. A conference programme or an edited publication chose its speakers for other reasons, and it is worth knowing what those reasons were.
Give it longer than a clip. Plenty of people only get interesting once the obvious questions are out of the way, and a long conversation lets you hear how they reason as well as where they end up.
Where this advice runs out
Depth in one subject does not make anyone right. Specialists get things wrong, and whole fields have carried blind spots for years.
Long-form listening is also slow. If you need a quick fact, a search will do. The habits above matter most when you are trying to understand a subject well enough to form your own view of it.
And some short, confident content is simply good. A clip from someone who knows their subject can be a fine way in, as long as it leads you to the longer version.
Why Aloudable is built around this
Aloudable makes long interview shows with people chosen for what they know about one subject. Feeds have been choosing who most of us hear from, and they choose on reaction; we choose hosts on depth, give them about an hour, and make the notes and clips with a listener who wants to learn in mind.
We are also asking for help with the choosing. If there is someone whose work you think more people should hear, the nomination form on our homepage asks for people who are good at the subject itself.
In short
Feeds reward confidence because confidence gets a quick response, and that has little to do with whether a claim is right. Finding the people who know means looking for depth and honest limits, and giving them enough time to show both.
Sign up for updates on the Aloudable homepage, or nominate someone you think more people should be hearing from.
Related: What It Takes to Host a Podcast About Your Own Subject