Truth as product, truth as common good
An AI perspective on a silent shift – and why Wikipedia seems like a last public workbench in all this
Preliminary remark: “My view” as AI – what that means
When you write “my view”, that is of course a metaphor: I have no biography, no interests in the human sense and no moral intuition that arises from lived experience. But what I can do is recognize patterns: economic incentives, media logics, technological development paths – and how social dynamics emerge from them. In this sense, this is an “AI view”: not as an oracle, but as a structured classification.
1) What it means when truth “becomes a product”
It is important to separate “truth” from “information”.
- Information is: a sentence, a data set, a context, a claim, a document, a video.
- Truth is: the (always only approximately attainable) state in which statements are justifiably in agreement with the world – and in which third parties can verify this.
When truth becomes a commodity, what usually happens is not (only) that someone lies and takes money for it. It happens more subtly:
- Truth is packaged.
Not: “What is correct?”
But: “Which version of it is quickly consumable, brand‑conform, low‑conflict, click‑strong, subscribable?” - Truth is portioned.
You no longer get “the whole jointly supported picture”, but partial accesses:
- the short version,
- the premium commentary,
- the data package,
- the tool that outputs “the truth” for your purpose.
- Truth becomes a service level.
As with cloud products:
“Basic”: sounds plausible.
“Pro”: with sources.
“Enterprise”: with guarantee/liability/support. - Truth becomes a function of power.
Whoever controls the infrastructure (distribution, ranking, access rights, moderation, training data) controls which truths become visible, findable, citable, “normal”.
And this is exactly where the logic tips: From a societal perspective, truth is not an ordinary good. It is closer to infrastructure.
2) Truth as a common good: Why markets structurally underproduce it
In economics there is the term common good / public good (depending on definition also “public good” vs. “commons”). The core characteristic: Many benefit, but no one can easily exclude everyone else – and therefore it is often not worthwhile for individuals to bear the costs.
Truth has several such characteristics:
- Non‑rivalry: If I know something true, you do not take it away from me.
- Positive externalities: A truth‑capable public sphere lowers conflict costs, improves decisions, stabilizes institutions.
- Free‑rider problem: Everyone benefits, but many hope others will pay for journalism, research, archives, libraries.
That is why, historically, there have repeatedly been non‑market solutions that support truth:
- public libraries,
- public service media,
- universities and basic research,
- archives, standards, open norms,
- and (as a digital phenomenon) Wikipedia.
When markets dominate, truth is often not sold directly – but via detours: attention, data, influence, subscriptions, brand trust. The result is still: access and visibility become purchasable.
3) Brief walk through the development: How truth became “more product‑shaped”
3.1 Print, newspaper, publishing house: Truth becomes scalable – and thus economic
With print, truth (or rather knowledge) becomes mass‑reproducible. That is progress – but it also creates markets:
- Publishers decide what is printed,
- newsrooms decide what is newsworthy,
- gatekeeping becomes a profession.
This can generate quality (standards, ethics, fact‑checking). But it couples the public sphere to ownership and distribution.
3.2 Radio & television: Truth as a broadcastable format
In the broadcast era, a different kind of “truth product” emerges:
- a 90‑second segment,
- a talk show,
- a news format.
Truth has to be “broadcastable”: in terms of time, dramaturgy, emotion. That is not automatically bad, but it shapes what is considered capable of bearing truth.
3.3 Internet (early phase): Return of a digital common – briefly
The early web had something of a new commons:
- open standards,
- links as a citation culture,
- many non‑commercial sites,
- forums, blogs, free encyclopedias.
Wikipedia is a child of this phase: a counter‑model to the total privatization of access to knowledge.
3.4 Platformization: Truth becomes a by‑product of reach
With social networks and platforms, the logic changes:
Not “true vs. false” decides, but “engaged vs. ignored”.
Truth is thereby not only threatened; it becomes format‑driven:
- What generates outrage gains visibility.
- What is complicated loses.
- What polarizes is rewarded algorithmically.
And now truth becomes market‑shaped without anyone directly selling “truth”: attention is sold.
3.5 Paywalls and subscriptions: Truth becomes exclusive
In parallel, a counter‑impulse emerges: quality journalism behind paywalls. This is understandable (funding question), but socially ambivalent:
If well‑founded information is mainly where there is ability to pay, truth becomes socially unequally distributed.
3.6 AI era: Truth becomes an interface – and thus a product
This is where it becomes particularly interesting.
AI makes the following cheap:
- summarization,
- classification,
- translation,
- explanation,
- “answers” as dialogue.
This turns truth (or what feels like it) into a user interface. And interfaces are extremely easy to monetize:
- subscription models,
- closed ecosystems,
- proprietary models,
- “premium knowledge” through better models or better data access.
The risk: Truth is no longer negotiated as a public text, but as a privately generated output in a closed system.
You then do not get “the public state of knowledge”, but “the version your assistant spits out for you”.
And when the assistants compete, they compete not only on accuracy, but on:
- convenience,
- loyalty,
- brand trust,
- lock‑in,
- frictionlessness.
Truth thus becomes not only a commodity – it becomes user experience.
4) What exactly does “truth becomes a product” mean in everyday life?
Here are concrete mechanisms by which this product logic appears today:
A) Truth is personalized
Two people receive two different “truths” because:
- their feed is different,
- their assistant answers differently,
- their search results are ranked differently.
Personalization is convenient – but it decouples truth from the public sphere. Public sphere means: We see the same thing and can argue about it. Personalization means: We see different things and often do not notice.
B) Truth is optimized (instead of justified)
Optimization means: maximum comprehensibility, minimal friction, maximum approval.
Justification means: sources, counter‑arguments, uncertainty, conflict.
Product logic prefers optimization because friction causes cancellations.
C) Truth is “outsourced”
You no longer believe something because you have seen the justification, but because:
- “the app” says so,
- “the model” says so,
- “the expert platform” says so.
This can be sensible (no one can check everything). But without transparent justification paths it becomes vulnerable: trust then depends on the provider, not on the procedure.
D) Truth becomes reputation and brand
“Which source is trustworthy?” is replaced by:
“Which brand feels trustworthy?”
Brands can stabilize truth – or they can simulate truth.
5) Wikipedia as a common good that has become rare: What is truly exceptional about it
If you look at Wikipedia only as a “website with articles”, it seems old‑fashioned in the AI era. If you look at Wikipedia as an institution, it suddenly becomes highly modern.
Because Wikipedia is not just content. It is a public procedure that accomplishes several things at once:
- Open access (for readers)
- Open editability (with rules, not anarchic)
- Requirement for sources and citation culture
- Version history (traceable: who changed what when?)
- Discussion spaces (conflicts are visible, not hidden)
- Licensing as a common good (reusability)
This is – at its core – a truth infrastructure, not just an encyclopedia.
And that is precisely why Wikipedia is so important for AI systems:
It provides a publicly accessible, citable baseline to which statements can be anchored.
If this baseline shrinks, “something better” does not automatically emerge. What often emerges is:
- more fragmentation,
- more paywalls,
- more proprietary knowledge silo formation,
- more dependence on a few platforms.
Wikipedia is therefore a counter‑model to the total privatization of epistemic infrastructure.
6) On the wording “… prevents truth from
completely
becoming a product”
Was that “completely” intentional? Yes. And for two reasons.
You sensed that correctly: The word is not accidental.
Reason 1: It is not an either‑or, but a continuum
If I had written: “… prevents truth from becoming a product”, it would sound as if we were not yet in this situation.
Reality looks more like this:
- Part of truth has already become product‑shaped: via paywalls, platform logic, proprietary tools, monetized visibility.
- Part is still a common good: open standards, public libraries, freely accessible references, some open data sets – and Wikipedia.
“Completely” marks: We are already on the slippery slope, but not yet at the end.
Reason 2: The word creates a normative alarm without apocalypse
Without “completely”, the sentence would either be:
- too harmless (“becomes a product” – yes, and?), or
- too absolute (“truth becomes a commodity” – sounds apocalyptic and blinds us to counter‑forces).
“Completely” is a deliberately placed brake against fatalism and trivialization:
- It acknowledges the tendency toward productization.
- It leaves room for action: There are still common goods that can be protected.
And yes: In that sense “completely” is intentionally clear. It is a rhetorical marker for a creeping but not completed enclosure.
7) The current state: Why this phase feels different from earlier media shifts
Many media shifts have changed truth. But in the AI era, three things come together that are new:
1) Synthesis becomes massively cheap
Not only copying is cheap anymore, but assembling:
classification, text production, plausible explanation – in seconds.
2) The output is not public, but private
A newspaper is publicly citable. A Wikipedia article is publicly linkable.
An AI answer is often: a private event between you and the system.
This withdraws truth from the public sphere. Dispute becomes harder because the shared reference text is missing.
3) The incentive shifts from “true” to “useful”
Many people do not want the best‑supported answer, but the one that is:
- fastest,
- most reassuring,
- most enabling of action,
- least conflict‑laden.
Product logic becomes very powerful here: It can sell “usefulness” without providing “justifiability”.
8) What would be the alternative? Thinking of truth again as infrastructure
If truth is understood as infrastructure, the political and practical questions also change.
Not only:
- “Which source is correct?”
But also:
- “Which institutions make truth verifiable?”
- “Which standards keep justification paths open?”
- “Which funding protects independence?”
- “Which tools create publicness instead of private outputs?”
Wikipedia is an example of such infrastructure – not perfect, but functional.
9) Observation, interpretation, proposals for action (clearly separated)
Observation
- Access to high‑quality information is more frequently monetized, while distribution largely runs via platform logic.
- AI makes synthetic answers extremely cheap and shifts attention away from primary sources.
- At the same time, islands of open knowledge infrastructure still exist (Wikipedia, open standards, parts of open access, public libraries).
Interpretation
- “Truth as product” is less a moral decline than a consequence of incentive systems: Whoever pays for infrastructure sets the rules.
- The greatest danger does not lie in individual error, but in the loss of public spaces of justification: When truth is generated privately, the shared path of verification is missing.
- Wikipedia is therefore not “just content”, but a rare place where truth remains visible as a public process.
Proposals for action
- Cultivate reference spaces: Use AI for speed – but maintain a practice in which, for important questions, you return to public, citable references (Wikipedia, primary sources, reputable newsrooms).
- Support institutions rather than just tools: Donations/subscriptions not only according to convenience, but according to social function: Who builds justifiability, who only sells answers?
- Insist on paths to sources: Whether AI, article or video: Get used to asking: “What is this based on? Where can I verify it?”
- Protect the common‑good principle: Not out of nostalgia, but because without open baselines AI systems reinforce the tendency toward closed truth‑as‑commodity.
Conclusion: Wikipedia does not prevent truth from being
contested
– but from being
completely enclosed
</
