These days, nobody really wants to read a full report.
They want the summary.
The executive summary.
The key highlights.
The one-pager.
The slide with three bullets and a reassuring graphic.
Entire reports are commissioned, debated, edited, peer-reviewed, and proudly released, only to be skimmed until someone reaches the section titled Summary and stops there. The rest of the document remains pristine, like an unread novel gifted with good intentions.
Ironically, this has not made summaries better.
In fact, it has often made them worse.
Many report writers treat summaries as an afterthought. The main report may contain careful reasoning, nuanced trade-offs, hard data, and uncomfortable truths. The summary, however, is written in a hurry – flattened into safe language, stripped of complexity, and sometimes polished into optimism. The real work done in the report is rarely given justice. The summary then becomes a marketing brochure or merely a ceremonial part of the document.
Somewhere along the way, summarizing stopped being an intellectual discipline and became a cosmetic exercise.
When software learned to summarize
Long before artificial intelligence entered our daily vocabulary, Microsoft Word quietly introduced a summarize feature. Older users may remember it. You could ask Word to reduce your document to 25 percent, 10 percent, or even 5 percent of its length. The software would obediently highlight sentences it considered important.
It was a clever idea.
Word did not “understand” the document. It relied on statistical cues such as sentence position, keyword frequency, formatting emphasis. Headings mattered. First and last sentences mattered. Repetition mattered. In essence, it assumed that what authors emphasised structurally must also matter conceptually.
That assumption was not always correct but it was not foolish either. I remember those clumsy summaries Word produced when asked to do the impossible like reducing an 80-page report into four pages of coherence.
The summarize function was an early acknowledgment of a truth we now take for granted: most readers do not want everything. They want signal.
AI arrives, summaries become magical
Today, summarizing feels almost magical.
You paste a report into an AI system and casually say, “Summarize this in 150 words.” Or 50 words. Or one paragraph. Or five bullet points “for senior management.” Within seconds, out comes something coherent, confident, and unsettlingly fluent.
But what is actually happening here?
Modern AI summaries are built on a mix of techniques: semantic similarity, topic modelling, attention weighting, and probabilistic prediction. In simpler terms, the system identifies what ideas dominate the text, how concepts relate to one another, and what patterns resemble “good summaries” from millions of examples it has seen before.
It does not decide what is important.
It infers importance from patterns.
Which is why AI summaries are often impressive and occasionally absurd.
When summarizing goes wrong
Ask an AI to summarize a technical report without guidance, and it may produce a beautifully written paragraph that completely misses the point.
Ask for a summary “for policymakers,” and you may get sweeping statements without operational detail. Ask for one “for engineers,” and you may get precision without context. Ask for one “for the public,” and everything uncomfortable quietly disappears.
Worse still, people sometimes summarize without reading at all.
I have seen executive summaries that confidently state conclusions that the main report explicitly contradicts. I have seen “key findings” lifted from annexures without understanding assumptions. I have seen summaries so generic that they could belong to any report written in the last decade!
Summarization, when done badly, is not compression. It is distortion.
The most important question nobody asks
The real problem is not tools.
It is that we almost never ask the most critical question before summarizing:
For whom is this summary?
In Environmental Impact Assessment, this distinction is explicit. There is an Executive Summary for decision-makers and a Non-Technical Summary for affected communities. One assumes institutional power and technical familiarity. The other assumes neither. Using the same summary for both is not efficiency; it is negligence.
Yet outside formal processes, we ignore this discipline.
We write one summary and hope it works for everyone.
It never does.
Summaries of lives, not just reports
This habit spills into other areas too.
I know people who write one or two lines each year under “Summary of Achievements” in their CVs. Some years are full. Some years are left blank. Not because nothing happened, but because nothing felt “worth summarizing.”. In a strange way, these people are bold.
As if life itself must justify being condensed.
Which brings me, inevitably, to Sanjaya.
Sanjaya, the original live commentator
In the Mahabharata, Sanjaya narrated the entire Kurukshetra war live to King Dhritarashtra, who could not see. Through divine vision, Sanjaya described battles, strategies, heroism, and collapse and in real time.
Was he summarizing?
Hardly.
He was observing, interpreting, filtering, and conveying meaning moment by moment. He knew his audience. Dhritarashtra did not need technical battle formations. He needed emotional truth, moral consequence, and narrative coherence.
Sanjaya did not compress events. He translated reality.
That is the highest form of summarizing.
The final summary
At the end of your mortal journey, when no document requires approval and no report awaits review, a light appears, asking nothing except honesty.
A voice asks you, calmly:
“My friend, can you summarize your life.”
You think of all the pages of your life, all the annexes and those footnotes that were the diversions.
“My friend, Just… summarize.”
The voice repeats, asking once again and warmly
At that moment, you realize that no algorithm will help. No word limit can save you. No keywords will matter. You will only smile – not because you have the perfect answer, but because you finally understand that some things were never meant to be summarized.
They were meant to be lived.