Viewing enriched Twitter post
Yes! The “are you sure?” problem (link below) is especially pervasive in any complex coding task. Ask Claude or GPT to review a PR then ask it to double check its findings when it finishes - it’ll flip on at least 1 of its findings. https://t.co/WUSuOxTuDy
{
"media": [
{
"url": "https://crmoxkoizveukayfjuyo.supabase.co/storage/v1/object/public/media/posts/2033928292101984370/media_0.jpg",
"media_url": "https://crmoxkoizveukayfjuyo.supabase.co/storage/v1/object/public/media/posts/2033928292101984370/media_0.jpg",
"type": "photo",
"filename": "media_0.jpg"
}
],
"processed_at": "2026-03-17T15:36:48.126845",
"pipeline_version": "2.0"
}{
"type": "tweet",
"id": "2033928292101984370",
"url": "https://x.com/randal_olson/status/2033928292101984370",
"twitterUrl": "https://twitter.com/randal_olson/status/2033928292101984370",
"text": "Yes! The “are you sure?” problem (link below) is especially pervasive in any complex coding task. Ask Claude or GPT to review a PR then ask it to double check its findings when it finishes - it’ll flip on at least 1 of its findings.\n\nhttps://t.co/WUSuOxTuDy",
"source": "Twitter for iPhone",
"retweetCount": 1,
"replyCount": 1,
"likeCount": 0,
"quoteCount": 0,
"viewCount": 312,
"createdAt": "Tue Mar 17 15:27:48 +0000 2026",
"lang": "en",
"bookmarkCount": 0,
"isReply": false,
"inReplyToId": null,
"conversationId": "2033928292101984370",
"displayTextRange": [
0,
257
],
"inReplyToUserId": null,
"inReplyToUsername": null,
"author": {
"type": "user",
"userName": "randal_olson",
"url": "https://x.com/randal_olson",
"twitterUrl": "https://twitter.com/randal_olson",
"id": "49413866",
"name": "Randy Olson",
"isVerified": false,
"isBlueVerified": true,
"verifiedType": null,
"profilePicture": "https://pbs.twimg.com/profile_images/1996406365125005312/LjQz6KGT_normal.jpg",
"coverPicture": "https://pbs.twimg.com/profile_banners/49413866/1764815390",
"description": "Co-Founder & CTO @GoodeyeLabs. Building tools that put domain experts in control of AI quality. AI & ML researcher. Data viz aficionado. 15+ years building AI.",
"location": "Pacific Northwest USA",
"followers": 111470,
"following": 27675,
"status": "",
"canDm": true,
"canMediaTag": true,
"createdAt": "Sun Jun 21 20:13:18 +0000 2009",
"entities": {
"description": {
"urls": []
},
"url": {
"urls": [
{
"display_url": "randalolson.com",
"expanded_url": "https://randalolson.com",
"indices": [
0,
23
],
"url": "https://t.co/C7FtnCU9fH"
}
]
}
},
"fastFollowersCount": 0,
"favouritesCount": 33485,
"hasCustomTimelines": false,
"isTranslator": false,
"mediaCount": 5994,
"statusesCount": 21219,
"withheldInCountries": [],
"affiliatesHighlightedLabel": {},
"possiblySensitive": false,
"pinnedTweetIds": [
"2029919935770636294"
],
"profile_bio": {},
"isAutomated": false,
"automatedBy": null
},
"extendedEntities": {},
"card": {
"binding_values": [
{
"key": "description",
"value": {
"string_value": "Ask your AI 'are you sure?' and watch it flip. Models fold 60% of the time because we trained them to please, not push back. The fix isn't better prompts.",
"type": "STRING"
}
},
{
"key": "domain",
"value": {
"string_value": "www.randalolson.com",
"type": "STRING"
}
},
{
"key": "photo_image_full_size_alt_text",
"value": {
"string_value": "The \"Are You Sure?\" Problem: Why Your AI Keeps Changing Its Mind",
"type": "STRING"
}
},
{
"key": "vanity_url",
"value": {
"scribe_key": "vanity_url",
"string_value": "randalolson.com",
"type": "STRING"
}
},
{
"key": "summary_photo_image_alt_text",
"value": {
"string_value": "The \"Are You Sure?\" Problem: Why Your AI Keeps Changing Its Mind",
"type": "STRING"
}
},
{
"key": "title",
"value": {
"string_value": "The \"Are You Sure?\" Problem: Why Your AI Keeps Changing Its Mind",
"type": "STRING"
}
},
{
"key": "card_url",
"value": {
"scribe_key": "card_url",
"string_value": "https://t.co/WUSuOxTuDy",
"type": "STRING"
}
}
],
"card_platform": {
"platform": {
"audience": {
"name": "production"
},
"device": {
"name": "Swift",
"version": "12"
}
}
},
"name": "summary_large_image",
"url": "https://t.co/WUSuOxTuDy",
"user_refs_results": []
},
"place": {},
"entities": {
"hashtags": [],
"symbols": [],
"timestamps": [],
"urls": [
{
"display_url": "randalolson.com/2026/02/07/the…",
"expanded_url": "https://randalolson.com/2026/02/07/the-are-you-sure-problem-why-your-ai-keeps-changing-its-mind/",
"indices": [
234,
257
],
"url": "https://t.co/WUSuOxTuDy"
}
],
"user_mentions": []
},
"quoted_tweet": {
"type": "tweet",
"id": "2033920254938886331",
"url": "https://x.com/HamelHusain/status/2033920254938886331",
"twitterUrl": "https://twitter.com/HamelHusain/status/2033920254938886331",
"text": "One thing that makes me feel that code factory has not arrived yet is the following experiment:\n\n1.Ask a LLM to do an in-depth rigorous review of your code\n2. In a new thread, as same/different LLM to consider those review comments independently and address issues it agrees with\n3. Keep repeating until no new concerns\n\nI find that this loop always goes on for a ridiculously long time, which means that there is a problem with the notion of claude-take-the-wheel. This seems to happen no matter the harness or the specificity of the specs. \n\nIt works fine for simple applications, but in the limit if the LLMs have this much cognitive dissonance you cannot trust it. \n\nEither this, or LLM are RLHFd to always find some kind of issue.",
"source": "Twitter for iPhone",
"retweetCount": 2,
"replyCount": 13,
"likeCount": 27,
"quoteCount": 1,
"viewCount": 1619,
"createdAt": "Tue Mar 17 14:55:51 +0000 2026",
"lang": "en",
"bookmarkCount": 11,
"isReply": false,
"inReplyToId": null,
"conversationId": "2033920254938886331",
"displayTextRange": [
0,
280
],
"inReplyToUserId": null,
"inReplyToUsername": null,
"author": {
"type": "user",
"userName": "HamelHusain",
"url": "https://x.com/HamelHusain",
"twitterUrl": "https://twitter.com/HamelHusain",
"id": "825766640",
"name": "Hamel Husain",
"isVerified": false,
"isBlueVerified": true,
"verifiedType": null,
"profilePicture": "https://pbs.twimg.com/profile_images/1287206199088173057/ixE4fKy1_normal.jpg",
"coverPicture": "https://pbs.twimg.com/profile_banners/825766640/1758993452",
"description": "Bringing data science back to AI - https://t.co/Zrmp6LRd9c\n\nAbout Me: https://t.co/P6WyeKkyTa",
"location": "Looking at the data",
"followers": 44274,
"following": 2452,
"status": "",
"canDm": true,
"canMediaTag": false,
"createdAt": "Sat Sep 15 18:45:02 +0000 2012",
"entities": {
"description": {
"urls": [
{
"display_url": "evals.info",
"expanded_url": "http://evals.info",
"indices": [
36,
59
],
"url": "https://t.co/Zrmp6LRd9c"
},
{
"display_url": "hamel.dev",
"expanded_url": "https://hamel.dev",
"indices": [
71,
94
],
"url": "https://t.co/P6WyeKkyTa"
}
]
},
"url": {
"urls": [
{
"display_url": "evals.info",
"expanded_url": "http://evals.info",
"indices": [
0,
23
],
"url": "https://t.co/Zrmp6LRd9c"
}
]
}
},
"fastFollowersCount": 0,
"favouritesCount": 17424,
"hasCustomTimelines": true,
"isTranslator": false,
"mediaCount": 1541,
"statusesCount": 16231,
"withheldInCountries": [],
"affiliatesHighlightedLabel": {},
"possiblySensitive": false,
"pinnedTweetIds": [
"1998452926935695649"
],
"profile_bio": {},
"isAutomated": false,
"automatedBy": null
},
"extendedEntities": {},
"card": null,
"place": {},
"entities": {
"hashtags": [],
"symbols": [],
"urls": [],
"user_mentions": []
},
"quoted_tweet": null,
"retweeted_tweet": null,
"article": null
},
"retweeted_tweet": null,
"article": null
}