๐Ÿฆ Twitter Post Details

Viewing enriched Twitter post

@rohanpaul_ai

Yann LeCun (@ylecun ) beautifully explains how the architecture and principles used to train LLMs can not be extended to teach AI the real-world intelligence. In 1 line: LLMs excel where intelligence equals sequence prediction over symbols. Real-world intelligence requires learned world models, abstraction, causality, and action planning under uncertainty, which current next-token training does not provide. He says current LLMs learn by predicting the next token. That objective works very well when the task itself can be reduced to manipulating discrete symbols and sequences. Math, physics problem solving on paper, and coding fit this pattern because success largely comes from searching and composing the right sequences of symbols, equations, or program tokens. With enough data and scale, these models get very good at that kind of structured sequence prediction. Real-world intelligence is different. The physical world is continuous, noisy, uncertain, and high dimensional. To act in it, a system needs internal models that capture objects, dynamics, causality, constraints from the body, and the outcomes of actions over time. Humans and animals build abstract representations from rich sensory streams, then make predictions in that abstract space, not at the raw pixel level. That is why a child can learn intuitive physics, plan multi-step actions, and adapt quickly in new situations with little data. His claim about saturation follows from this gap. Scaling token prediction keeps improving symbol manipulation tasks like math and code, but it hits limits on embodied reasoning and common sense because text alone does not provide the right learning signals for world models. Predicting the next word cannot efficiently teach contact forces, affordances, occlusion, friction, or how actions change the state of the environment. For that, he argues we need architectures that learn abstractions from sensory data and predict futures in abstract latent spaces, then use those predictions to plan actions toward goals with built-in guardrails. --- From 'Pioneer Works' YT Channel (link in comment)

๐Ÿ“Š Media Metadata

{
  "media": [
    {
      "url": "https://crmoxkoizveukayfjuyo.supabase.co/storage/v1/object/public/media/posts/2001673936300576892/media_0.mp4?",
      "media_url": "https://crmoxkoizveukayfjuyo.supabase.co/storage/v1/object/public/media/posts/2001673936300576892/media_0.mp4?",
      "type": "video",
      "filename": "media_0.mp4"
    }
  ],
  "processed_at": "2025-12-19T01:22:40.040918",
  "pipeline_version": "2.0"
}

๐Ÿ”ง Raw API Response

{
  "type": "tweet",
  "id": "2001673936300576892",
  "url": "https://x.com/rohanpaul_ai/status/2001673936300576892",
  "twitterUrl": "https://twitter.com/rohanpaul_ai/status/2001673936300576892",
  "text": "Yann LeCun (@ylecun ) beautifully explains how the architecture and principles used to train LLMs can not be extended to teach AI the real-world intelligence.\n\nIn 1 line: LLMs excel where intelligence equals sequence prediction over symbols. Real-world intelligence requires learned world models, abstraction, causality, and action planning under uncertainty, which current next-token training does not provide.\n\nHe says current LLMs learn by predicting the next token. That objective works very well when the task itself can be reduced to manipulating discrete symbols and sequences. Math, physics problem solving on paper, and coding fit this pattern because success largely comes from searching and composing the right sequences of symbols, equations, or program tokens. With enough data and scale, these models get very good at that kind of structured sequence prediction.\n\nReal-world intelligence is different. The physical world is continuous, noisy, uncertain, and high dimensional. To act in it, a system needs internal models that capture objects, dynamics, causality, constraints from the body, and the outcomes of actions over time. Humans and animals build abstract representations from rich sensory streams, then make predictions in that abstract space, not at the raw pixel level. That is why a child can learn intuitive physics, plan multi-step actions, and adapt quickly in new situations with little data.\n\nHis claim about saturation follows from this gap. Scaling token prediction keeps improving symbol manipulation tasks like math and code, but it hits limits on embodied reasoning and common sense because text alone does not provide the right learning signals for world models. Predicting the next word cannot efficiently teach contact forces, affordances, occlusion, friction, or how actions change the state of the environment. For that, he argues we need architectures that learn abstractions from sensory data and predict futures in abstract latent spaces, then use those predictions to plan actions toward goals with built-in guardrails.\n\n---\n\nFrom 'Pioneer Works' YT Channel (link in comment)",
  "source": "Twitter for iPhone",
  "retweetCount": 16,
  "replyCount": 20,
  "likeCount": 104,
  "quoteCount": 4,
  "viewCount": 15701,
  "createdAt": "Thu Dec 18 15:20:30 +0000 2025",
  "lang": "en",
  "bookmarkCount": 78,
  "isReply": false,
  "inReplyToId": null,
  "conversationId": "2001673936300576892",
  "displayTextRange": [
    0,
    275
  ],
  "inReplyToUserId": null,
  "inReplyToUsername": null,
  "author": {
    "type": "user",
    "userName": "rohanpaul_ai",
    "url": "https://x.com/rohanpaul_ai",
    "twitterUrl": "https://twitter.com/rohanpaul_ai",
    "id": "2588345408",
    "name": "Rohan Paul",
    "isVerified": false,
    "isBlueVerified": true,
    "verifiedType": null,
    "profilePicture": "https://pbs.twimg.com/profile_images/1816185267037859840/Fd18CH0v_normal.jpg",
    "coverPicture": "https://pbs.twimg.com/profile_banners/2588345408/1729559315",
    "description": "Compiling in real-time, the race towards AGI.\n\nThe Largest Show on X for AI.\n\n๐Ÿ—ž๏ธ Get my daily AI analysis newsletter to your email  ๐Ÿ‘‰ https://t.co/6LBxO8215l",
    "location": "Ex Inv Banking (Deutsche)",
    "followers": 116062,
    "following": 8355,
    "status": "",
    "canDm": true,
    "canMediaTag": false,
    "createdAt": "Wed Jun 25 22:38:54 +0000 2014",
    "entities": {
      "description": {
        "urls": [
          {
            "display_url": "rohan-paul.com",
            "expanded_url": "https://www.rohan-paul.com",
            "url": "https://t.co/6LBxO8215l",
            "indices": [
              134,
              157
            ]
          }
        ]
      },
      "url": {
        "urls": [
          {
            "display_url": "rohan-paul.com",
            "expanded_url": "http://www.rohan-paul.com",
            "url": "https://t.co/2NKnK0wIil",
            "indices": [
              0,
              23
            ]
          }
        ]
      }
    },
    "fastFollowersCount": 0,
    "favouritesCount": 54918,
    "hasCustomTimelines": true,
    "isTranslator": false,
    "mediaCount": 25358,
    "statusesCount": 62544,
    "withheldInCountries": [],
    "affiliatesHighlightedLabel": {},
    "possiblySensitive": false,
    "pinnedTweetIds": [
      "1965551636082032917"
    ],
    "profile_bio": {
      "description": "Compiling in real-time, the race towards AGI.\n\nThe Largest Show on X for AI.\n\n๐Ÿ—ž๏ธ Get my daily AI analysis newsletter to your email  ๐Ÿ‘‰ https://t.co/6LBxO8215l"
    },
    "isAutomated": false,
    "automatedBy": null
  },
  "extendedEntities": {
    "media": [
      {
        "display_url": "pic.x.com/2153m7OPSX",
        "expanded_url": "https://x.com/rohanpaul_ai/status/2001673936300576892/video/1",
        "id_str": "2001673141136084992",
        "indices": [
          276,
          299
        ],
        "media_key": "13_2001673141136084992",
        "media_url_https": "https://pbs.twimg.com/amplify_video_thumb/2001673141136084992/img/w4O8yWMCP-NeTTcN.jpg",
        "type": "video",
        "url": "https://t.co/2153m7OPSX",
        "additional_media_info": {
          "monetizable": false
        },
        "ext_media_availability": {
          "status": "Available"
        },
        "sizes": {
          "large": {
            "h": 1080,
            "w": 1920,
            "resize": "fit"
          },
          "medium": {
            "h": 675,
            "w": 1200,
            "resize": "fit"
          },
          "small": {
            "h": 383,
            "w": 680,
            "resize": "fit"
          },
          "thumb": {
            "h": 150,
            "w": 150,
            "resize": "crop"
          }
        },
        "original_info": {
          "height": 1080,
          "width": 1920,
          "focus_rects": []
        },
        "allow_download_status": {
          "allow_download": true
        },
        "video_info": {
          "aspect_ratio": [
            16,
            9
          ],
          "duration_millis": 73216,
          "variants": [
            {
              "content_type": "application/x-mpegURL",
              "url": "https://video.twimg.com/amplify_video/2001673141136084992/pl/BKq_uq4kfKj3xZqb.m3u8?v=c1e"
            },
            {
              "bitrate": 256000,
              "content_type": "video/mp4",
              "url": "https://video.twimg.com/amplify_video/2001673141136084992/vid/avc1/480x270/yenBdnGXQYoa9r3h.mp4"
            },
            {
              "bitrate": 832000,
              "content_type": "video/mp4",
              "url": "https://video.twimg.com/amplify_video/2001673141136084992/vid/avc1/640x360/VRIRS8dY44lOr92M.mp4"
            },
            {
              "bitrate": 2176000,
              "content_type": "video/mp4",
              "url": "https://video.twimg.com/amplify_video/2001673141136084992/vid/avc1/1280x720/pPEBSMaFof058uS-.mp4"
            },
            {
              "bitrate": 10368000,
              "content_type": "video/mp4",
              "url": "https://video.twimg.com/amplify_video/2001673141136084992/vid/avc1/1920x1080/v0EvoynPh-dKeZsx.mp4"
            }
          ]
        },
        "media_results": {
          "result": {
            "media_key": "13_2001673141136084992"
          }
        }
      }
    ]
  },
  "card": null,
  "place": {},
  "entities": {
    "hashtags": [],
    "symbols": [],
    "timestamps": [],
    "urls": [],
    "user_mentions": [
      {
        "id_str": "48008938",
        "name": "Yann LeCun",
        "screen_name": "ylecun",
        "indices": [
          12,
          19
        ]
      }
    ]
  },
  "quoted_tweet": {
    "type": "tweet",
    "id": "2001322361514336541",
    "url": "https://x.com/rohanpaul_ai/status/2001322361514336541",
    "twitterUrl": "https://twitter.com/rohanpaul_ai/status/2001322361514336541",
    "text": "Yann LeCun's new interview - explains why LLMs are so limited in terms of real-world intelligence.\n\nSays the biggest LLM is trained on about 30 trillion words, which is roughly 10 to the power 14 bytes of text. \nThat sounds huge, but a 4 year old who has been awake about 16,000 hours has also taken in about 10 to the power 14 bytes through the eyes alone. So a small child has already seen as much raw data as the largest LLM has read.\n\nBut the childโ€™s data is visual, continuous, noisy, and tied to actions: gravity, objects falling, hands grabbing, people moving, cause and effect. From this, the child builds an internal โ€œworld modelโ€ and intuitive physics, and can learn new tasks like loading a dishwasher from a handful of demonstrations.\n\nLLMs only see disconnected text and are trained just to predict the next token. So they get very good at symbol patterns, exams, and code, but they lack grounded physical understanding, real common sense, and efficient learning from a few messy real-world experiences.\n\n---\n\nFrom 'Pioneer Works' YT channel (link in comment)",
    "source": "Twitter for iPhone",
    "retweetCount": 80,
    "replyCount": 53,
    "likeCount": 485,
    "quoteCount": 21,
    "viewCount": 96217,
    "createdAt": "Wed Dec 17 16:03:28 +0000 2025",
    "lang": "en",
    "bookmarkCount": 298,
    "isReply": false,
    "inReplyToId": null,
    "conversationId": "2001322361514336541",
    "displayTextRange": [
      0,
      279
    ],
    "inReplyToUserId": null,
    "inReplyToUsername": null,
    "author": {
      "type": "user",
      "userName": "rohanpaul_ai",
      "url": "https://x.com/rohanpaul_ai",
      "twitterUrl": "https://twitter.com/rohanpaul_ai",
      "id": "2588345408",
      "name": "Rohan Paul",
      "isVerified": false,
      "isBlueVerified": true,
      "verifiedType": null,
      "profilePicture": "https://pbs.twimg.com/profile_images/1816185267037859840/Fd18CH0v_normal.jpg",
      "coverPicture": "https://pbs.twimg.com/profile_banners/2588345408/1729559315",
      "description": "Compiling in real-time, the race towards AGI.\n\nThe Largest Show on X for AI.\n\n๐Ÿ—ž๏ธ Get my daily AI analysis newsletter to your email  ๐Ÿ‘‰ https://t.co/6LBxO8215l",
      "location": "Ex Inv Banking (Deutsche)",
      "followers": 116062,
      "following": 8355,
      "status": "",
      "canDm": true,
      "canMediaTag": false,
      "createdAt": "Wed Jun 25 22:38:54 +0000 2014",
      "entities": {
        "description": {
          "urls": [
            {
              "display_url": "rohan-paul.com",
              "expanded_url": "https://www.rohan-paul.com",
              "url": "https://t.co/6LBxO8215l",
              "indices": [
                134,
                157
              ]
            }
          ]
        },
        "url": {
          "urls": [
            {
              "display_url": "rohan-paul.com",
              "expanded_url": "http://www.rohan-paul.com",
              "url": "https://t.co/2NKnK0wIil",
              "indices": [
                0,
                23
              ]
            }
          ]
        }
      },
      "fastFollowersCount": 0,
      "favouritesCount": 54918,
      "hasCustomTimelines": true,
      "isTranslator": false,
      "mediaCount": 25358,
      "statusesCount": 62544,
      "withheldInCountries": [],
      "affiliatesHighlightedLabel": {},
      "possiblySensitive": false,
      "pinnedTweetIds": [
        "1965551636082032917"
      ],
      "profile_bio": {
        "description": "Compiling in real-time, the race towards AGI.\n\nThe Largest Show on X for AI.\n\n๐Ÿ—ž๏ธ Get my daily AI analysis newsletter to your email  ๐Ÿ‘‰ https://t.co/6LBxO8215l"
      },
      "isAutomated": false,
      "automatedBy": null
    },
    "extendedEntities": {
      "media": [
        {
          "display_url": "pic.x.com/ip5gZVmx4E",
          "expanded_url": "https://x.com/rohanpaul_ai/status/2001322361514336541/video/1",
          "id_str": "2001321049603100672",
          "indices": [
            280,
            303
          ],
          "media_key": "13_2001321049603100672",
          "media_url_https": "https://pbs.twimg.com/amplify_video_thumb/2001321049603100672/img/N4zu3ZFcQZVJQXD5.jpg",
          "type": "video",
          "url": "https://t.co/ip5gZVmx4E",
          "additional_media_info": {
            "monetizable": false
          },
          "ext_media_availability": {
            "status": "Available"
          },
          "sizes": {
            "large": {
              "h": 1080,
              "w": 1920,
              "resize": "fit"
            },
            "medium": {
              "h": 675,
              "w": 1200,
              "resize": "fit"
            },
            "small": {
              "h": 383,
              "w": 680,
              "resize": "fit"
            },
            "thumb": {
              "h": 150,
              "w": 150,
              "resize": "crop"
            }
          },
          "original_info": {
            "height": 1080,
            "width": 1920,
            "focus_rects": []
          },
          "allow_download_status": {
            "allow_download": true
          },
          "video_info": {
            "aspect_ratio": [
              16,
              9
            ],
            "duration_millis": 53994,
            "variants": [
              {
                "content_type": "application/x-mpegURL",
                "url": "https://video.twimg.com/amplify_video/2001321049603100672/pl/0GuOyKZevFs_FW6V.m3u8?v=581"
              },
              {
                "bitrate": 256000,
                "content_type": "video/mp4",
                "url": "https://video.twimg.com/amplify_video/2001321049603100672/vid/avc1/480x270/GqJgDDoRQ33Gcwlj.mp4"
              },
              {
                "bitrate": 832000,
                "content_type": "video/mp4",
                "url": "https://video.twimg.com/amplify_video/2001321049603100672/vid/avc1/640x360/FTHjvEaEWhnRnVPH.mp4"
              },
              {
                "bitrate": 2176000,
                "content_type": "video/mp4",
                "url": "https://video.twimg.com/amplify_video/2001321049603100672/vid/avc1/1280x720/KR6_NIGIJY_SK6aR.mp4"
              },
              {
                "bitrate": 10368000,
                "content_type": "video/mp4",
                "url": "https://video.twimg.com/amplify_video/2001321049603100672/vid/avc1/1920x1080/ePMECpnS5Yr6UHvJ.mp4"
              }
            ]
          },
          "media_results": {
            "result": {
              "media_key": "13_2001321049603100672"
            }
          }
        }
      ]
    },
    "card": null,
    "place": {},
    "entities": {
      "hashtags": [],
      "symbols": [],
      "timestamps": [],
      "urls": [],
      "user_mentions": []
    },
    "quoted_tweet": null,
    "retweeted_tweet": null,
    "article": null
  },
  "retweeted_tweet": null,
  "article": null
}