Wheeling, West Virginia Correspondent — They amplified good information and bad information at enormous scale, and the word “both” in this question is doing more work than usual.

Holding them accountable for every piece of content is unworkable and turns them into censors. Holding them accountable for systemic amplification patterns — particularly paid or coordinated distribution of demonstrably false health claims — is reasonable.

What I’d note from a place with low trust in institutions to begin with is that the removals didn’t work the way anyone intended. Taking down a video here didn’t make people believe the official version. It made them certain the video had been correct.

Prescott Valley, Arizona Correspondent — Amplification is the core issue. Platforms design the systems deciding what millions see first, and design is a choice even when nobody makes it deliberately.

Accountability should focus on those design choices and on coordinated falsehood rather than on liability for individual posts.

I’d add an observation from teaching. What the year actually demonstrated is that most people have no framework for holding a claim provisionally — for saying this is the current best estimate and it may change. We don’t teach that, and a population without it will read every revision as a lie.

That’s an education failure that arrived disguised as a platform problem.

Gastonia, North Carolina Correspondent — They amplified everything at huge scale. More accountability for the systems doing the pushing is reasonable. Turning them into full-time referees of scientific disputes is not.

Here’s the part that ought to trouble everybody, and I say this as somebody who fact-checks copy for a living. Several claims that were removed as misinformation in the spring turned out to be defensible by the autumn. The lab-origin question is the obvious one.

If your enforcement mechanism can’t survive the science moving, it isn’t a fact-checking system. It’s a system for enforcing the current official position, which is a different product entirely and should be labeled as such.

Sheffield, Jamaica Correspondent — Systemic amplification of false health claims deserves scrutiny. Post-level liability does not.

I would raise the export problem again, since it applies here with unusual force. The health information policies these firms adopted were written against an American argument and applied globally.

In my region that meant enforcement against claims about treatments and remedies calibrated to a debate happening in the United States, applied to populations with different health systems, different drug availability, and a different relationship with official medicine. Nobody consulted anyone here.

Knoxville, Tennessee Correspondent — They amplified official statements that aged poorly and suppressed unofficial statements that sometimes aged well. That isn’t a science desk. That’s an editorial desk in a safety costume.

Accountability means a log. What was boosted, what was suppressed, and which official claims later broke.

Hold them to consistency. If they claim to fight falsehood they apply the standard to the podium as well as to the crank, and this year the podium was wrong repeatedly and never got a label.

The remedy is sunlight on the mechanism, not a second authority telling them to suppress more.

Long Island, New York Correspondent — Accountability is process, not a promise to be correct. Nobody was correct all year, including the people doing the enforcing.

So what I’d require is what I’d require of any party with discretionary authority: publish the rule, publish the exception list, and keep a record when you apply it.

No quiet reduction of a physician’s reach because the mood shifted. If a claim is being suppressed, somebody signs for it and it goes in a file that can be read afterward.

That doesn’t guarantee good judgment. It guarantees that bad judgment is reviewable, which is the most any procedure can offer.

Jacksonville, Florida Correspondent — A longshoreman watched official advice reverse and unofficial video disappear, and he drew a conclusion about the referee. He wasn’t being irrational. He was reasoning from what he could see.

A referee that wants trust shows the tape. Show what you amplified. Live with the embarrassment, because embarrassment is cheaper than a public that assumes every removal is a plot.

Some removals were mistakes and some were incompetence and a few were probably right. A log sorts them. Nothing else does.

Las Vegas, Nevada Correspondent — Posted odds that favor the house are fine. Odds that change without notice lose you the room, and that’s what happened.

Accountable means posted rules, applied the same way, and a willingness to say “we boosted this and it was wrong.”

Without that last sentence they aren’t public-health partners. They’re participants holding a mute button, and a mute button requires a record.