Skip to content
The Future Explorerby Benjamin Talin
Enablertechnological

Social Networks

Services whose main job is connecting people to each other, such as Facebook, Instagram, TikTok and YouTube, paid for by the attention those connections hold. They change how friendships, couples and opinions form. About 3.6 billion people use one company's apps every single day. Over the same decade, the share of Americans who spend part of a typical day socialising in person or by phone fell from 38% to 30%. Whether the feeds caused that is still disputed.

Maturity

5/ 5
measured
BasisSaturated: 3.6 billion people use Meta's apps every day, and 84% of US adults use YouTube. The constraint has moved to rules - minimum ages, platform duties, forced sales. The effects literature is contested.SourceE-168 (BLS ATUS); E-075 (GSS via IFS); Meta, Q2 2026 results (29 Jul 2026); Pew Research Center (20 Nov 2025); BLS ATUS 2025 (25 Jun 2026); Prime Minister of Australia (16 Jan 2026); European Commission (5 Dec 2025); NASEM (Dec 2023)Vintage2023–2026
ReadinessCapability 5Deployment 5Absorption 5Bound byregulationTrenddeteriorating

Where it feeds in

5 megatrends in 3 forces

Where you have already met this

In 2015, 38% of Americans spent some part of an average day socialising and communicating. In 2025, 30%. Time on games and leisure computing rose from 25 minutes a day to 37. The feed is the obvious suspect. When the US National Academies weighed the evidence on teenagers, it did not support a population-level verdict. Suspicion is not a finding.

If this moved

If this entry fell to 3 - through rules that changed design for everyone, not only minors - platform mediation of social life would share its constraint. Nothing else would move: the relationship recession, the attention economy, polarisation and epistemic fragmentation are held lower by family formation, generational turnover, consumption norms and epistemic trust.

What to watch for

What the European Commission does next under the Digital Services Act after its first fine, €120 million against X in December 2025. Whether Australia's under-16 rule reduces teenagers' use or moves it elsewhere.

Not to be confused with

Messaging apps. A private chat goes to people you chose. A social network ranks content and pushes it to people who did not ask for it.

What it is

Services whose primary function is mediating relationships between people, and whose economics depend on the time those relationships consume.

How it works

  1. 01Map who knows whomThe platform keeps a graph: people are points; follows and friendships are the links between them.
  2. 02Record every reactionThe app logs what you stop on, watch to the end, like or share - down to how long you paused.
  3. 03Shortlist candidatesA fast model pulls a few hundred likely posts out of millions.
  4. 04Rank for engagementA larger model scores each candidate on how likely you are to react, and orders the feed.
  5. 05Sell the attentionAdvertising is placed in the feed. More time on the app means more ads shown.

State of play

  • Meta averaged 3.60 billion daily active people across its apps in June 2026, up 3% on a year earlier; ad impressions rose 14% (Meta, Q2 2026 results, 29 Jul 2026).
  • 84% of US adults use YouTube, 71% Facebook, 50% Instagram and 37% TikTok - TikTok up from 21% in 2021 (Pew Research Center, 20 Nov 2025).
  • The share of Americans socialising and communicating on an average day fell from 38% in 2015 to 30% in 2025 (BLS, American Time Use Survey, 25 Jun 2026).
  • Australia's minimum age of 16 for social media took effect on 10 December 2025; platforms deactivated, removed or restricted more than 4.7 million under-16 accounts (Prime Minister of Australia, 16 Jan 2026).
  • The European Commission's first non-compliance decision under the Digital Services Act fined X €120 million, on 5 December 2025 (European Commission, 5 Dec 2025).
  • The US Supreme Court upheld the law, signed on 24 April 2024, requiring TikTok to be sold or banned (TikTok v. Garland, 17 Jan 2025).

Why it matters

It restructures how relationships form and are maintained. That is a claim about the human enabler layer - family formation, adoption appetite and epistemic trust all run partly through this technology now.

  • It is saturated and feeds five megatrends, none of them held back by it. The technology works; the open questions are what it does to people and who governs it.
  • The causal case is unsettled. The National Academies' 2023 review did not support population-level harm to adolescent health; the fall in socialising is measured.
  • Rules are now where the gradient changes - minimum ages, platform duties, forced sales.

What holds it back

Age and design rules

Australia bars under-16 accounts; the EU requires the largest platforms to assess and reduce systemic risks.

Researcher access

Settling cause and effect needs platform data. The EU fined X partly for blocking researchers from it.

Ownership and national security

The US TikTok law made ownership, not features, the condition for operating.

Terms

Social graph
The network of people and the links between them that a platform keeps.
Engagement ranking
Ordering a feed by how likely each post is to make you react.
Dwell time
How long you pause on a post - a signal of interest the ranking uses.
Very large online platform (VLOP)
A service with more than 45 million EU users, which carries the Digital Services Act's heaviest duties.
Algorithmic feed
A feed ordered by a model's predictions rather than by time.

Steepens

The gradient behaves like a sticky rubber band: it is pulled hardest toward the weakest enabler. Marginal links are shown but do not pull.

If this moved one step

Scored 5 of 5: there is no step above it.

What if it moved?

What would change the gradient

Conditional, not predictive. No dates, no probabilities.

  1. 01Design defaults around minors, which is where regulation is actually moving
  2. 02Researcher access to platform data - the constraint on settling the causal question
  3. 03Alternative venues for the social functions displaced, which is not a technology problem

Common misconception

That the measured decline in in-person socialising is straightforwardly caused by it. The correlation is strong, the timing is suggestive, and the causal literature is genuinely unsettled. Say so.

Talin, B. (2026). Social Networks. The Future Explorer, dataset v0.4.0 (2026-11-16). https://explorer.talin.digital/e/social-networks/