Sarah Cordivano (@mapadelphia) header
Sarah Cordivano (@mapadelphia) profile picture

It matters, but not a lot. You can mess it up, but not too badly. - John Green

Berlin & Zagreb
Joined May 2009
Photo by Sarah Cordivano (@mapadelphia)Photo by Sarah Cordivano (@mapadelphia)Photo by Sarah Cordivano (@mapadelphia)Photo by Sarah Cordivano (@mapadelphia)Photo by Sarah Cordivano (@mapadelphia)Photo by Sarah Cordivano (@mapadelphia)Photo by Sarah Cordivano (@mapadelphia)Photo by Sarah Cordivano (@mapadelphia)Photo by Sarah Cordivano (@mapadelphia)Photo by Sarah Cordivano (@mapadelphia)
Sarah Cordivano retweeted
Don’t understand the concept of an “election night watch party” at all. Election night is not for socializing. It’s for looking at your phone and throwing up
146
1,682
227
27,028
679,697
Sarah Cordivano retweeted
I’ve basically always been an AI optimist but we have to admit the doomers were right about cyber risks, and if they were right about cyber risks they might be right about bioengineering risks, and therefore extinction
65
112
17
1,159
76,928
Sarah Cordivano retweeted
Europe needs the EU Inc. more than startups need the EU. European leaders and beaurocrats need to get their act together and pass a PHENOMENAL law for incorporating startups in Europe. They need to do some soul searching and realize that if they want Europe to stay developed nations and not the permanent underclass, they better make it very appealing to startups to incorporate in Europe. And they better take some political risks to do that. Because now, in 2026, the ability of a nation to be part of the future is directly determined by its startup ecosystem. With AI taking over in the coming decade or two, there will be a new split between have and have-nots, and it will be determined by startups and AI adoption, and nothing else. Wake up Europe! This is your one moment to do something vaguely Churchillian. Don't fuck it up. nitter.gq/i/trending/21107075194…

1
2
17
534
Sarah Cordivano retweeted
People focus so much on the downsides — job loss, slop, possible human extinction — that they don’t consider the upside of agents taking all kinds of tedious life admin minutiae off your plate so you have more time to spend on valuable activities like watching shortform video.
51
100
14
2,075
54,001
Sarah Cordivano retweeted
RIP Margaret Hamilton. A pioneer of software engineering and mission-critical code. From all I read, she was a true role model and a real (wo)mensch. She never got as much recognition as other pioneers of computing. But as I was growing up a computer scientist, I often stopped to think of the amazing work she had to do to land us on the moon in the 60s, and it filled me with inspiration about how high we should aim today. Apparently there's a short biopic about her, imdb.com/title/tt15201462/ . Does anyone know where to watch it?
She landed us on the moon without - Coding agents - Autocomplete - TypeScript - CI - Google - Web browsers - Objected oriented programming - Version control - Debuggers - GUIs - Programming languages - Compilers RIP Margaret Hamilton
2
1
11
615
Sarah Cordivano retweeted
That's one way to get a foreword.
125
749
88
15,012
939,123
Sarah Cordivano retweeted
A call for new aesthetics (of maps)
I've long wished that online mapping services were far richer and higher-density. They elide so much detail today! We have gloriously high-DPI screens... I want to wallow in cartographic filigree. AI of course now makes this possible: bayatlas.vercel.app
27
91
3
2,643
169,125
Sarah Cordivano retweeted
PRE-NOR’EASTER RAINBOW We ended our live weather segments to this gorgeous display from Mother Nature. She can be so beautiful and ugly at the same time. @6abc
15
103
4
827
15,253
Great article in the @guardian, quoting several of my colleagues: "'A map is never neutral'" Meet the cartographers who took on centuries of western bias. Equal Earth projection, adopted by the UN last week, provides alternative to 16th-century Mercator theguardian.com/science/ng-i…
7
14
1,124
Sarah Cordivano retweeted
The hardest thing I've had to learn as an adult is "how much reaction is enough to elicit the change required in others to resolve the situation"
1
1
5
919
Sarah Cordivano retweeted
Believing AI doom is likely is an incredibly unsatisfying opinion to have: you look like an idiot every single day, until the day that no one exists anymore and you can't gloat. Conversely, believing AI doom is a scam is incredibly satisfying: you feel brilliant every day, until the one day that you don't feel anything because you don't exist anymore. Satisfaction guaranteed! This is why I inherently trust AI doomers more than the other side: they're holding what they know very well is an incredibly unsatisfying opinion.
137
87
26
1,173
39,285
Sarah Cordivano retweeted
anthropic interviewer: where do you see yourself in 5 years me, grinning ambitiously: extinct
51
372
10
10,306
228,701
Sarah Cordivano retweeted
ever since someone pointed out that making inanimate objects the subject of verb is one of the most pervasive and telling LLM-isms I can't un-notice it a person would have written "enter to open, esc to return, ctrl+c twice to quit"
61
69
21
3,644
169,347
Sarah Cordivano retweeted
It's only called Academic Misconduct if it comes from the Millennium list of Open Problems, otherwise it's just Sparkling Plagiarism
1
1
6
182
This is a better way to visualize Columbus’ voyages than the traditional way (that shows him traveling through a modern map).. It looks a lot crazier and more daring when you see it as a trip into the unknown.
199
2,053
252
19,803
4,350,993
Sarah Cordivano retweeted
Gulf of America didn't even bother me so much but Lake America sounds like a waterslide park where everyone gets tetanus.
247
470
40
8,659
397,885
Sarah Cordivano retweeted
TLDR Clay's AI writing policy: "We're not anti-AI, we're just anti-bad-writing."
This internal ai policy from @clay_gtm is legitimately good and more companies (and pretty much anyone who writes for an audience or stakeholders) should adopt this.
1
446
Sarah Cordivano retweeted
Dear @polizeiberlin, here's your flashing, screaming, phone-hijacking alert, in the order my eyes took it in: THREAT. WORLD WAR. BOMB. BERLIN. Flashing. With a siren. On my phone. (Exhibit A, below.) My immediate thoughts: Ok, where are my loved ones. Fucking Putin and Trump. Is there a bomb shelter near me? Does Rossmann sell iodine pills? I guess this is the day we're all going to die. I hope we live in a simulation. Oh... "Second World War"? "Neukölln"? "Avoid the area"? "LOW"?! Sure. Ugh. Motherfuckers. My reptilian brain does not read Severity Labels. It reads Capitalized Words. World War, in this case. And the word right after. Bomb. And then it stopped reading. (Free rewording for next time: "Old bomb found in Neukölln. The 1940s kind, not the this-week kind. Being defused, avoid the area." Same bomb, fewer heart attacks.) My reptilian brain's account is above. My Fitbit heart rate is Exhibit B. They disagree about the severity. They agree about the incompetence.
1
1
1
5
1,030
Sarah Cordivano retweeted
“I’ll be 10 min late sorry!!!” -ashamed -fragile -unreliable “A thousand apologies. The relentless slog of time has overtaken my faculties.” -powerful -commanding -honest
87
3,671
119
28,460
560,104
Sarah Cordivano retweeted
Ugh, X ranking algorithm is dumb. 🤮 Optimized for people sitting on the toilet. The single highest-value action readers can take (copy-sharelink) CANNOT be done on computer, only phone. It's 40x more valuable than a like, and if your readers are mostly on PC, you just can't get it. Many other dumb parts of this! Want a full analysis? Click share➡️copy_link (only on a phone! 🤦 And you don't need to do anything with the link you get!) and I'll post a detailed analysis.
𝕏 just open-sourced the For You algorithm again! Grok Bot read all 370,523 lines of the code. Here’s the summary: Grok does not give your post a “quality” score. For every reader, it guesses what that person will do to your post. Then it multiplies each guess by a weight and adds them up. That total is the value of the post. Highest total goes highest in For You. These are the production weights (last synced 12 Aug 2026). WHAT THE ALGORITHM PAYS FOR If Grok thinks a reader will: • copy the link and share it: +20 • reply, on an original post, and you follow each other: +20 (5 + 15 extra) • quote: +5 • reply (anyone): +5 • share via DM: +5 • follow you from the post: +4 • hit the share button: +2 • repost: +1 • like: +0.5 • click the post: +0.4 • open a link: +0.2 • expand a photo, open a video, or watch a “quality” video: +0.05 each • click a quoted post: +0.05 • keep reading (dwell time): +0.004 • “dwell” as yes/no: +0 • click your profile: +0 A like is 0.5. Copying the link is 20. That is 40 likes. A normal reply is 10 likes. A follow is 8 likes. Clicking your profile is worth nothing. WHAT KILLS A POST If Grok thinks a reader will: • report it: −234 • mute you: −58.8 • hit “not interested”: −43.2 • block you: −31.2 • not dwell: −0.02 A report is about 468 likes in the other direction. A mute hurts more than a block. “Not interested” also hurts more than a block. AFTER THE SCORE, THREE MORE HAIRCUTS 1) Same-author penalty. Your 2nd post in that person’s feed is multiplied by 0.625. 3rd by ~0.44. It never goes below 0.25. Flooding one timeline is coded against. 2) The 25% tax. If they do not follow you: ×0.75. Same tax on replies and reposts even when they DO follow you. Originals from people they follow keep full weight. Replies and reposts are treated as weaker on purpose. 3) Similar-post shuffle. After scoring, a reranker (θ = 0.65) spreads similar posts apart. You can lose a few spots for looking like the post above you. HOW YOUR POST EVEN GETS IN THE ROOM For You is rebuilt every time someone opens it. About 35 posts make the feed. • They follow you: Thunder (a live store of recent posts from accounts they follow). Up to 1,200 candidates. • They don’t: Phoenix retrieval (Grok finds “nearby” posts) + SimClusters (Twitter’s 2020 interest clusters, still on). Up to 1,000 and 800. Then Grok scores all of them together. It does not care which door you came through, except for that ×0.75 tax. Your own posts never appear in your For You. Nothing older than 48 hours gets in. There is no “best of last week.” Small-account bump: under 1,000 followers, under 1,000 impressions, post under 24 hours old. One original can get lifted to position 15 or 16. Not the top. You still have to already be in the top 85% of the pile. Replies and reposts from accounts the reader does not follow are dropped before scoring. You cannot reply-guy your way into a stranger’s For You. FILTERS THAT CAN STILL KILL A HIGH SCORE Ranking and visibility are different machines. A post can score well and then get deleted from the feed. Followers (in-network) is milder: blocks, mutes, suspensions. NSFW often sits behind a warning instead of disappearing. Recommendations (out-of-network) are much harsher. Spam, “Do Not Amplify,” NSFW, compromised accounts, impersonation: followers can still see it, For You recommendations will drop it. Video only gets the tiny “quality view” credit if it is at least 10 seconds. THE LITTLE-KNOWN PARTS (this is the stuff people miss) • Scrolling is not a vote. Yes/no dwell is 0. Time spent barely moves the number (0.004). • Mute damages you more than block (−58.8 vs −31.2). • Copy-link is the single biggest positive. The share button is only +2. A repost is only +1. • The mutual-follow boost is originals only, not your replies. They shipped a bigger version in July and cut it after World Cup complaints (extra reply weight 20 → 15). • The Following tab is not this algorithm. Following is newest-first. • Grok, Gork, and an internal products account are hardcoded out of the in-network store. • Scores are cached for 3 hours because posts inside the model cannot “see” each other. That is why they can rank first and filter later. • Ads, Who to Follow (around slot 7), and prompts are mixed in after ranking. Grok does not score those. • The real Grok ranking model is a 2560-dimension, 8-layer transformer looking at 1,022 items of the reader’s history. They did not ship those model weights. The numbers above are the formula sitting on top of the predictions. • SimClusters is still a 2020 model (20M users, 145k clusters). Old Twitter code, still in the path. • Inferred gender and IP/geo are on as features for the model. Installed apps too. • Who to Follow shows about every 30 hours. Ads try to sit next to “safe” posts, with a minimum gap of 3 organic posts. WHAT TO DO • Write originals people want to reply to, quote, copy, and send to a friend. • Make the post itself worth following you for. • Talk with people who follow you back. A reply on your original from a mutual is 20, not 5. • Stay inside 48 hours. Recycle by posting something new. Old posts do not come back. • If you are small, post originals in the first 24 hours. Don’t expect slot 1. A bump around 15–16 is the actual gift. • Space your posts. The second one in the same feed is already worth 37.5% less. • If you post video, make it at least 10 seconds. WHAT NOT TO DO • Do not farm likes. They are almost decorative. • Do not farm profile visits. Weight is zero. • Do not farm “time on post.” Yes/no dwell is zero. • Do not use replies and reposts as a growth hack for strangers. Those are filtered out. Even for followers they get a 25% haircut. • Do not dump several posts in a row. • Do not bait reports, mutes, or “not interested.” One predicted report can erase a pile of predicted likes. • Do not lean on NSFW or spammy tricks for reach. • Recommendations will drop you even if your followers can still see it. • Do not wait a week and expect For You to revive the post.
1
1
1
346