17 connected korrents · 16 moments on record from 21 Jan 2019 to 1 Sept 2026.
At the centre
Language models have codified formal written language at a moment in time, which leaves informal, inventive, in-group speech as the place where humanness still shows.
Last stated 31 Dec 2022 · 4 years ago
Holds MAMaggie Appleton
Read this korrent →
-
Similar wording
Language models have no culture — no written record they hand to one another — and that absence is one of the things holding them back.
Last stated 17 Oct 2025 · 11 months ago
Holds
Andrej Karpathy
-
Similar wording
Editing your writing with a model pulls it towards one homogeneous style, far enough that you lose track of what your own voice was.
Last stated 19 Aug 2026 · 3 weeks ago
Holds AOAddy Osmani
-
Similar wording
Formal verification is about to become economical, because models are getting good enough at writing the proofs that humans no longer have to.
Last stated 22 Apr 2026 · 5 months ago
Holds MKMartin Kleppmann
-
Similar wording
Language models are already a form of AGI; what the labs are chasing is a further step, not the arrival of general intelligence.
Last stated 3 Feb 2025 · 2 years ago
Holds NLNathan Lambert
-
Similar wording
A language should be designed by a small group of people rather than by one person.
Last stated 13 May 2026 · 4 months ago
Holds AHAnders Hejlsberg
-
Similar wording
Large language models do not reason formally: their performance collapses as a problem is made bigger, in the way a calculator's never does.
Last stated 11 Oct 2024 · 2 years ago
Holds GMGary Marcus
-
Similar wording
AI-written prose is an inferior read because a model emits a statistical average -- code's audience is a machine, writing's audience is people.
Last stated 1 Sept 2026 · a week ago
Holds
Gergely Orosz
-
Similar wording
Letting more people write software in natural language does not reduce the work; it snowballs the amount of software that has to be maintained.
Last stated 22 Mar 2025 · a year ago
Holds THThePrimeagen
-
Same subject: formal proof
Formalising a proof in Lean currently takes about ten times the effort of writing it out: doable, but annoying.
Last stated 14 Jun 2025 · a year ago
Holds TTTerence Tao
-
Same subject: formal proof
Full formal verification is a waste of money for most software: near-perfect is reachable with ordinary techniques at a fraction of the cost.
Last stated 21 Jan 2019 · 8 years ago
Holds HWHillel Wayne
-
Same subject: formal proof
Lean and tools like GitHub will let experimental mathematics scale far beyond what one mathematician's spaghetti code allows today.
Last stated 14 Jun 2025 · a year ago
Holds TTTerence Tao
-
Same subject: AI and human skill
AI can be a shortcut around thinking or a way of thinking more rigorously, and for your core job function only the second one is wanted.
Last stated 12 Aug 2026 · 4 weeks ago
Holds CMCharity Majors
-
Same subject: AI and human skill
AI should be treated as an instrument to play rather than a tool to use, because there is no purpose to better machines if they do not also produce better humans.
Last stated 20 Oct 2025 · 11 months ago
Holds FCFrank Chimero
-
Same subject: AI and human skill
Heavy AI use erodes understanding first and judgement second: cognitive debt turns into cognitive surrender, where the model's answer simply becomes your answer.
Last stated 19 Aug 2026 · 3 weeks ago
Holds AOAddy Osmani
-
Same subject: AI slop
A growing fraction of social media is AI slop influencers being replied to by bots — an echo of an echo of an echo.
Last stated 23 Aug 2026 · 2 weeks ago
Holds
François Chollet
-
Same subject: AI slop
A low-effort AI-designed landing page costs a product its audience before the product itself is ever judged.
Last stated 13 Aug 2026 · 4 weeks ago
Holds
Mitchell Hashimoto
-
Same subject: AI slop
AI-generated writing tends to repeat the same themes, names, and underlying ideas across different outputs.
Last stated 31 Aug 2026 · a week ago
Holds EMEthan Mollick