Who coined the term information overload?

Bertram Gross used it in 1964 in The Managing of Organizations, but Alvin Toffler is the one who spread it widely in 1970 with Future Shock. French later coined its own word for it, infobésité, which entered ordinary professional use in the 2000s.

Does information overload have measurable effects?

Yes. Gloria Mark’s work at the University of California, Irvine shows average attention on a screen fell from roughly 2.5 minutes in 2004 to 47 seconds in 2021, and that it takes more than 20 minutes on average to recover focus after an interruption.

Is cutting the number of sources enough?

It helps, but it is not enough. Overload comes as much from unprocessed material as from volume: ten undigested sources weigh more than fifty that were summarised and filed. The real lever is processing throughput, not intake.

What is information overload?

Information overload is the state in which the volume of incoming information exceeds a person’s capacity to process it, degrading their understanding and their decisions. Alvin Toffler popularised the term in 1970 in Future Shock.

What exactly are we talking about?

Information overload occurs when the amount of information to be processed exceeds the cognitive capacity available to process it. The symptom is not a shortage of information but having too much of it to draw anything from: decision quality degrades even as available material increases.

The phenomenon was identified well before the internet. Bertram Gross used the phrase information overload in 1964, and Alvin Toffler put it into circulation in 1970 in Future Shock, where he described people’s growing inability to absorb the rate of change around them. French coined "infobésité" in the 2000s, a term now standard in professional usage there.

What effects are documented?

The measurements converge on three effects, all counter-intuitive to anyone who assumes more information beats less.

Fragmented attention. Gloria Mark’s work at UC Irvine establishes that average attention on a screen dropped from about 2.5 minutes in 2004 to 47 seconds in 2021. Each interruption costs more than twenty minutes before focus returns to its prior level.
Worse decisions. Past a certain point, adding information stops improving the quality of a decision and instead raises the confidence of the person making it — a gap well documented since Paul Slovic’s 1973 work with horse-racing handicappers.
Choice paralysis. The jam study run by Sheena Iyengar and Mark Lepper in 2000 remains the canonical illustration: a display of 24 varieties draws more passers-by than one with 6, but produces ten times fewer purchases.

Why does the volume keep rising?

Mechanism	Effect on load
Near-zero cost of publishing	Supply grows with no natural ceiling
Algorithmic recommendation	The feed refills without the user acting
Long formats (podcast, video)	Ingestion time grows faster than volume
Notifications on by default	Load is imposed rather than chosen
Machine-generated content	Signal-to-noise ratio degrades

None of these mechanisms is going to reverse. That is what makes strategies built purely on voluntary reduction ineffective: they pit individual discipline against structural pressure.

Which strategies actually work?

Move from push to pull. Turn off notifications, replace algorithmic feeds with explicit subscriptions — RSS, channels, lists. The trigger to read becomes a decision again rather than an alert.
Decouple capture from reading. Save now, read later, in a dedicated slot. This avoids the interruption without losing the resource — provided the slot genuinely exists.
Filter on use, not interest. The question is not "is this interesting" — almost everything is — but "what will I do with it". That test eliminates 80 to 90 percent of candidates.
Summarise as a rule. An unsummarised source is still pending and still weighs on you. A summarised, filed source is out of the queue.
Delete without guilt. A queue nobody has touched in two months will not be touched. Emptying it is maintenance, not failure.

Why do capture tools often make it worse?

A fast capture tool — a browser extension, a read-it-later app — drops the cost of saving to one click. Intake rises immediately, while processing throughput stays capped by human reading time. The mechanical result is a queue that grows.

That is the read-it-later paradox: Pocket, founded in 2007 and shut down by Mozilla in 2025, ended up as an article graveyard for a great many of its users. The problem was not the tool but the asymmetry it installed between two steps, only one of which had been accelerated.

The same asymmetry explains why adding a second capture tool never helps. Two inboxes fill twice as fast and are read at the same rate as one. Any intervention on the intake side changes the shape of the backlog without changing its size.

How do you rebalance capture and processing?

The only durable way out is to accelerate processing as much as capture. That means automating what is mechanical: transcribing audio and video, producing a faithful synthesis, extracting the concepts, flagging overlaps with what has already been read, keeping the exact quotation with its source.

Synthiz is built on that symmetry: a source you add — article, video, podcast, PDF, RSS feed — comes back transcribed, synthesised and filed into a thematic "memory", its concepts linked to earlier reading. What enters the system leaves it processed, which is what stops the queue from forming.