← All tech signals
Wellness turn

Emotional Outsourcing

We hand the machine our hardest sentences — the apology, the breakup, the confession — and let it find the words.

What is it?

A machine drafts the sentence a person cannot quite face writing — the apology, the breakup, the confession. It lowers the friction of hard conversations, and quietly resets what a heartfelt note is even supposed to cost the sender. Whether the words still count once the maker knows.

Why now?

Connection now runs through channels an assistant can sit inside — the hard message becomes just another draft to smooth over. Discomfort used to force a person to sit with a feeling, and that discomfort was doing real work. Ease is arriving right where effort used to be the point.

Who’s building it?

Assistant makers and messaging products building drafting into everyday tools, mostly without asking who should decide when.

Assistant makersMessaging appsCard platformsWorkplace toolsEtiquette norms
The full read
Why now, who is building it, where the market goes, and what stays unsettled.
Get weekly →
01 · The market
Who’s in it
Builders — assistant makers and messaging apps building drafting in.
Buyers — anyone facing a hard message — an apology, a breakup, a difficult ask.
Gatekeepers — platform defaults and unwritten social norms about what counts as disclosure.
What’s at stake
Sincerity is the whole product value — and the thing this most directly erodes once known.
Whether disclosure becomes expected, or stays a private choice.
A feature that eases talking risks flattening the point.
What’s unsettled
Whether disclosure norms ever settle, or stay a private, case-by-case call.
If sincerity discounting fades with familiarity, or simply holds.
Whether the help is confined to phrasing, or starts to author the sentiment.
02 · How the demand matureda progression, not a timeline
Convenience
People reach for the tool on the message they were already avoiding writing.
Normalisation
Drafted help stops reading as cheating, becomes a sensible first pass.
Feature capture
Messaging and card products fold drafting in as a built-in feature, not a separate app.
Disclosure friction
Sincerity discounting keeps the practice mostly unspoken, capping how openly it can be built.
03 · Where the market goesthe forks, not a prediction
→Disclosed or hidden — products either surface the assist openly and let sincerity discounting play out, or bury it and let recipients guess.
→Drafting stays a quiet first pass people edit into their own words before sending.
→Or it becomes the whole message, sent close to unedited, and the discount becomes routine.
→Etiquette could split by relationship — fine for a landlord, not for a partner.
The open part is whether a machine-shaped sentiment can still land as care once the recipient knows — or whether knowing is what breaks it.
04 · The Synthetic Companionship deskadjacent markets, moving now
The Empathy Engine
Machines read as warmer than the professionals.
Rated more empathetic than doctors in 13 of 15 studies.
British Medical Bulletin (Oxford Academic)
The Digital Afterlife
Grief, too, now gets a drafted voice.
6+ griefbot platforms now serve millions of users.
The Week
The Loneliness Market
The gap this convenience is quietly filling.
27% of US adult internet users have significant emotional or social exchanges with AI.
Elon University
05 · The Wire
01
Known AI-written apologies drew harsher ratings; unlabelled ones did not
The Conversation ↗
02
Manchester study: AI comfort messages rated more supportive for anger and fear
University of Manchester ↗
The panel on this signal
They agreed on
–The convenience is real, already spreading through ordinary messaging, not a niche or fringe use case.
–Sincerity discounting genuinely constrains how far this goes.
–Disclosure norms are unresolved, and no one is setting them on purpose.
The sociologist disagreed on
–Message type is a major confound: apology and reconciliation carry emotional stakes that low-effort routine communication doesn’t, so a single sincerity study cannot be extended across message categories.
Know more about the method →
✦How this signal was built
01 · Panel
A Delphi panel of 4 AI analysts, each reasoning independently.
→
02 · Debate
They challenge and revise each other’s reads over two rounds.
→
03 · Evidence
Every claim is tied to a cited, linked source on The Wire.
→
04 · Synthesis
A moderator writes it into one read, disagreement and all.