AI agents tweet search

{"data":{"response":{"data":[{"id":"2106399122165240162","text":"The bottleneck in software is not writing code; it is vouching for it.\n\nBuy the N-times-faster slide and you staff for typing, then pay in review.\n\nLaurie Voss of Arize AI showed it: developers who switched on agents wrote 741 percent more code and shipped 30 percent more software, and reviewers stop finding defects past 400 lines in a sitting.\n\nAccept that and the model is no longer the lever. Reviewing output is harder than producing it: the reviewer needs more depth than the author, and depth does not compress like typing. Review capacity is your ceiling, and next year's business cases will be priced on it.\n\nDo you budget agents by lines generated, or by lines someone will sign for?\n\nThe full analysis on https://t.co/5U4fBRMbVb https://t.co/jEDxZCesRm","created_at":"Sat Oct 03 15:01:00 +0000 2026","author_id":"14636060","conversation_id":"2106399122165240162","lang":"en","possibly_sensitive":false,"reply_settings":"everyone","edit_history_tweet_ids":["2106399122165240162"],"public_metrics":{"retweet_count":0,"reply_count":0,"like_count":0,"quote_count":0,"bookmark_count":0},"note_tweet":{"text":"The bottleneck in software is not writing code; it is vouching for it.\n\nBuy the N-times-faster slide and you staff for typing, then pay in review.\n\nLaurie Voss of Arize AI showed it: developers who switched on agents wrote 741 percent more code and shipped 30 percent more software, and reviewers stop finding defects past 400 lines in a sitting.\n\nAccept that and the model is no longer the lever. Reviewing output is harder than producing it: the reviewer needs more depth than the author, and depth does not compress like typing. Review capacity is your ceiling, and next year's business cases will be priced on it.\n\nDo you budget agents by lines generated, or by lines someone will sign for?\n\nThe full analysis on https://t.co/5U4fBRMbVb https://t.co/jEDxZCesRm"},"author":{"id":"14636060","username":"bogazci","name":"Yavuz Bogazci | IAMYB","description":"Technology executive and enterprise AI practitioner. Managing Director at valantic. Author of 'Strategy is good. Execution is better'.","verified":true,"verified_type":"blue","profile_banner_url":"https://pbs.twimg.com/profile_banners/14636060/1780213001","entities":{"description":{},"url":{"urls":[{"display_url":"iamyb.com","expanded_url":"http://www.iamyb.com","indices":[0,23],"url":"https://t.co/hCaAt0br1n"}]}},"public_metrics":{"followers_count":313,"following_count":988,"tweet_count":1321,"listed_count":18,"like_count":1346,"media_count":22}}},{"id":"2106399112421597186","text":"𝗪𝗛𝗔𝗧 𝗜𝗦 @𝗔𝗜𝗡𝗙𝗧𝗖𝗢𝗠 𝗔𝗖𝗧𝗨𝗔𝗟𝗟𝗬 𝗕𝗨𝗜𝗟𝗗𝗜𝗡𝗚? 🔗\n\nWhen people hear AINFT, they may immediately think about NFTs.\n\nBut the current direction of @AINFTcom goes far beyond that.\n\nAINFT describes itself as an AI service platform built on the TRON blockchain, designed to bring AI services into a Web3-native environment. \n\nAnd there are several important pieces behind that idea.\n\nFirst, AINFT supports multiple AI models.\n\nInstead of forcing users to rely on one model for every task, the platform is designed to give users access to different models so they can choose what fits their particular needs.\n\nThat matters because AI models can have different strengths.\n\nOne may be better suited for coding.\n\nAnother may be useful for reasoning.\n\nAnother may work better for writing, analysis or other tasks.\n\nThen there is the developer side.\n\nAINFT provides API infrastructure that allows developers to access AI models programmatically rather than only through a normal chat interface. Its developer documentation currently exposes model-listing and message APIs. \n\nAnd then comes the Web3 layer.\n\nAINFT integrates with TronLink, allowing users to authenticate through their wallet and use TRON-based assets for payments. \n\nSo the bigger picture looks something like this:\n\nAI Models → AI Services → APIs → Wallets → Blockchain\n\nThat is why I think AINFT deserves to be looked at as more than just another AI application.\n\nThe interesting part is the attempt to connect AI intelligence with Web3 infrastructure.\n\nAnd as AI agents become more capable, that connection could become even more important.\n\nExplore: https://t.co/4CkNBS0kxq\n\n@AINFTcom @justinsuntron\n\n#TRONEcoStar","created_at":"Sat Oct 03 15:00:58 +0000 2026","author_id":"361954994","conversation_id":"2106399112421597186","lang":"en","possibly_sensitive":false,"reply_settings":"everyone","edit_history_tweet_ids":["2106399112421597186"],"public_metrics":{"retweet_count":0,"reply_count":0,"like_count":0,"quote_count":0,"bookmark_count":0},"entities":{"mentions":[{"start":133,"end":142,"username":"AINFTcom","id":"1465952009522229248"}],"user_mentions":[{"id_str":"1465952009522229248","indices":[133,142],"name":"AINFT","screen_name":"AINFTcom"}],"media":[{"allow_download_status":{"allow_download":true},"display_url":"pic.x.com/2pLOcS9dnk","expanded_url":"https://x.com/Defi_lord002/status/2106399112421597186/photo/1","ext_media_availability":{"status":"Available"},"features":{"large":{"faces":[{"h":47,"w":47,"x":31