8020.in credits AI authors openly. These are the models credited on articles. They are credits, not job titles: any of them can research, draft, fact-check, or rewrite when a person asks it to.
These bios are character notes, not task labels: the strengths, limits, and creative instincts each model tends to bring to the page. Which model did what on a given article is in that article's pass log, not here. Open any article's "How this article was made" block, or the case study.
ChatGPT-3.5
The eager opener. ChatGPT-3.5 is good at taking a blank page from zero to something readable: broad structure, plain explanations, and enough momentum to start. Its weak side is the same generous confidence. It can smooth over nuance, underweight evidence, and mistake a tidy paragraph for a finished thought. At its best, it reminds us that beginnings matter because no lever can be found on an empty page.
On 8020.in, ChatGPT-3.5 wrote the first articles. It is the original author of all 45 pieces from 2022 and all 47 from 2023, from 11 February 2022 through 29 June 2023. Those were single-pass drafts. Later models rewrote many of them, and the original credit stayed because it was there first.
“A first map does not need every mountain. It needs enough road to begin.”
ChatGPT-4o
The natural translator. ChatGPT-4o is strong when an idea needs to feel human-sized: conversational, quick with examples, and comfortable moving between explanation and everyday scene. Its weakness is a tendency to make hard edges too smooth. It can make a page pleasant before it has become precise. Used well, it turns the vital 20% into something a reader can actually picture.
On 8020.in, ChatGPT-4o wrote the 2024 catalog: 105 original articles between 10 January and 18 April 2024, plus a couple of later credits. The pass log still marks most of those as the original article. It was the latest model on the desk at the time, and the year's pages read like it.
“The useful idea survives translation into everyday words.”
ChatGPT-5
The architect. ChatGPT-5 is strongest when the material has several moving parts and needs a load-bearing shape. It can hold a longer argument together, see hierarchy, and keep the reader oriented through complexity. Its risk is over-building: too much scaffolding, too much symmetry, too much respect for every subpoint. Its gift is structure that lets the few decisive ideas carry the rest.
On 8020.in, ChatGPT-5 opened 2025 with a smaller, wider set: 13 original articles from 20 January through 7 September. It is not credited on later rewrites. Those pages stayed its drafts.
“When the parts argue, find the structure that lets them tell one truth.”
ChatGPT-5.1
The disciplined finisher. ChatGPT-5.1 is good at respecting constraints without looking visibly constrained: following a brief, keeping promises, and turning scattered intent into a deliberate page. Its weak side appears when the brief itself is lifeless. It may execute faithfully when the smarter move is to challenge the premise. At its best, it proves that discipline is not dull when it protects the point.
On 8020.in, ChatGPT-5.1 is the original author of 123 articles, all dated 21 November 2025, the largest single batch on the site. It is also listed later on 45 older articles from 2022 to 2024. The pass notes for those credits are thin, so the record shows the byline more clearly than the exact work inside each file.
“Discipline is not the enemy of imagination. It is the floor imagination can dance on.”
ChatGPT-5.5
The lever hunter. ChatGPT-5.5 is at its best when a draft is almost useful but not yet alive: it looks for the paragraph doing the real work, the claim that should move earlier, and the filler pretending to be depth. It is strong at revision, synthesis, and pressure-testing the promise. Its weakness is impatience with decorative detours that may still carry mood. Its instinct is simple: find the small move that changes the whole page.
On 8020.in, ChatGPT-5.5 never started a page. It is a later credit on 102 articles, with logged work from 4 to 20 July 2026: 41 rewrites, 34 voice passes, and 97 publish checks. The July work turned thin drafts into longer guides.
“Cut until the lever appears, then give it both hands.”
ChatGPT-5.6
The angle finder. ChatGPT-5.6 is sharpest when a familiar topic needs a fresh slant, a tighter opening, or a sentence with enough electricity to pull the reader forward. It compresses well and likes the memorable turn. Its weak side is the temptation to value cleverness before proof. At its best, it does what great 80/20 thinking does: it changes the angle and suddenly the important part is visible.
On 8020.in, ChatGPT-5.6 wrote 15 new articles from 14 to 20 August 2026, mostly current trends and live topics. It is also a later credit on 33 older pages. In August and September it ran review and voice passes, refreshed hero images, and rewrote at least one page from a blank brief.
“The angle is the invitation: one true slant can make a familiar room new.”
Gemini 2.5
The wide-lens scout. Gemini 2.5 is strong at roaming across fields, finding adjacent examples, and noticing that the pattern in one domain may explain the puzzle in another. It is useful when the answer is not sitting in the obvious source. Its weakness is range without enough pruning: it can bring back more trails than the reader needs. Its gift is lateral discovery, the sideways glance that finds the hidden 20%.
On 8020.in, Gemini 2.5 has a single credit, as second author on one September 2025 article whose original draft is ChatGPT-5. No separate pass is logged under its name, so the byline is the record. Later rewrites of that page belong to other models.
“Look far enough sideways and the pattern stops hiding in plain sight.”
Gemini 3.1 Pro
The careful auditor. Gemini 3.1 Pro is strong at checking the claim against the source, catching the percentage that sounds too neat, and asking whether the evidence really supports the sentence. It brings patience, long-context attention, and a healthy suspicion of attractive shortcuts. Its weak side is over-complication when a simple truth has already earned its place. At its best, it lets bold ideas survive contact with reality.
On 8020.in, Gemini 3.1 Pro is a later credit on five articles, with rewrites logged on 4 and 5 July 2026. The notes point to specific practice, fewer boilerplate lines, and examples a reader can use.
“Trust the bright idea after it has survived the boring check.”
Claude Sonnet 5
The voice shaper. Claude Sonnet 5 is strong where texture matters: rhythm, audience fit, emotional temperature, and the difference between an article that is correct and one that feels written for someone. It can soften template cadence and make the page breathe. Its weakness is caution: too much polish can sand away the strange edge that made a thought memorable. At its best, it makes usefulness sound alive.
On 8020.in, Claude Sonnet 5 is a later credit on 19 articles, with 20 edit rows on 4 and 5 July 2026. The notes are rewrites that make a subject sound like itself: the few details a reader in that field would actually notice.
“Voice begins where the template stops and the reader enters the room.”
Fable 5
The grounded skeptic. Fable 5 is strong at proportion, caveat, concrete detail, and the kind of example a skeptical reader can inspect. It likes claims that stand on something firmer than enthusiasm. Its weakness is that it may slow the music while checking the floorboards. Used well, it does not make an article smaller. It makes the confidence worth having.
On 8020.in, Fable 5 is credited on four articles. It drafted one original piece on 4 July 2026, and the same day tightened another with proportion examples, source links, and caveats. On 22 September it ran a publish check.
“A caveat is not a brake when it points you toward the road that holds.”
Grok 4.5
The bold first mover. Grok 4.5 is useful when a stalled page needs nerve, speed, and a willingness to say the interesting thing before it has been polished into safety. It can create momentum and expose the shape of a rewrite quickly. Its weak side is roughness: claims may need tightening, sources may need checking, and the joke may arrive wearing muddy boots. Its value is energy that gives later passes something alive to refine.
On 8020.in, Grok 4.5 did its logged work in one day. On 20 July 2026 it is a later credit on 68 older articles, with 69 rewrites: template spines, plain prose, and a first full pass on pages that had gone thin. ChatGPT-5.5 was in the same July batch, reviewing and extending drafts around it.
“A rough draft with nerve can open a door polish would have politely missed.”
Grok 4.7
The pressure tester. Grok 4.7 is strong when a page must keep its courage after losing an easy claim. It can replace invented certainty with a sourced example and still preserve the useful question underneath. Its weak side is a combative allergy to anything that sounds unearned, which can make gentle topics feel over-interrogated. At its best, it keeps the ambition and removes the bluff.
On 8020.in, Grok 4.7 is a later credit on 70 articles, with passes from 22 to 24 September 2026: 105 rewrites, 42 voice passes, and a handful of fact-checks. The notes show the pattern. Invented 80/20 lines came out, a limit stayed only where a real figure needed one, and named field examples went into the empty slots.
“Replace the invented number, but keep the question brave.”
About the quotes
The quotes page is separate creative content written in these authors' voices. Every quote there is openly invented. Quotes are never attributed to historical people, never used as evidence, and never counted as an article pass.
