Why I repeatedly emphasize that VLM (multimodal) is a dead end, and the world model is complete bullshit.
Summary
The author argues that VLMs (multimodal models) are a dead end because text has far greater information density than images or videos, and pure text LLM Agents can solve more complex problems.
View Cached Full Text
Cached at: 08/18/26, 06:25 AM
Why I repeatedly emphasize that VLMs (multimodal, vision-language models) are a dead end, and that world models are complete nonsense.
A simple first-principles argument: the information content of 100k tokens of text far exceeds that of images, and images vastly surpass video. A sufficient amount of text can solve some extraordinarily complex, complete problems. In contrast, image inputs to a VLM are extremely sparse, and the problems they can solve are essentially negligible.
If a VLM Agent, operating within a 1T to 10T parameter scale, can solve a problem of difficulty X, then an equivalent pure-text LLM Agent—when given ample context, memory, wiki, and all information prepared in pure text and structured data form, and when retrieving information via text/bash/cli/api/mcp interfaces—would undoubtedly be able to solve a problem of difficulty 1000X.
By the time VLMs become powerful enough for a VLM Agent to solve a problem of difficulty 1000X, an LLM Agent will already be tackling problems of difficulty (1000²) * X.
It’s that straightforward a principle, yet the vast majority of people simply fail to grasp it.
Similar Articles
@owmio39659: Introducing an important new paper and concept in the "jailbreak" field — "Babel Language": 1) LLMs don't actually need human-readable language to know what to do next. 2) Therefore, prompts and jailbreak content can be extremely compressed via an LLM into a segment of text unreadable to humans. 3) These unreadable texts can't be understood by humans, but the model can…
The paper proposes the concept of BabelTele ("Babel Language"), demonstrating that prompts and jailbreak content can be extremely compressed into text that is unreadable to humans yet retains recoverable semantics for LLMs — maintaining 99.5% semantic fidelity at only 27.9% of the original size. Applications include model offense/defense, multi-agent communication, and reducing context overhead.
@seclink: https://x.com/seclink/status/2067968283492712846
This article, based on the sharing of researcher Victoria Lin, systematically reviews the mainstream technical approaches of native multimodal large models (Chameleon, Transfusion, MOT) and their pros and cons. It points out that multimodal AI is still in the early exploration stage, with open problems such as gaps in scaling laws, inconsistency between image understanding and generation encoding, and connection with the physical world.
@Potatoloogs: How LLMs Actually Work Inside: From Token to Next-Token – A Complete Overview of Nine Core Mechanisms a) Tokenization: The model doesn't read text, it reads integers · Text is first split into subword pieces, then mapped to integer IDs; modern LLM vocabularies typically have tens of thousands to...
This article systematically outlines the nine core mechanisms inside modern LLMs, from tokenization to next-token prediction, including tokenization, embedding, positional encoding, attention, multi-head attention, feed-forward networks, etc., and compares architectural differences between various models.
@omarsar0: 1M-token, multimodal decision models are here! But there is more to this one. You can classify, extract, and summarize …
ScaleDown 推出 1M-token 多模态 Constrained Decoding Models (CDMs),支持文本与图像输入,用于分类、抽取和摘要任务,通过一次读取输入直接定位答案而非逐 token 生成,声称比前沿 LLM API 快 200 倍、便宜 467 倍。
@freeman1266: You don't need math to understand most AI papers—just understand this chain: token → embedding → position encoding → attention → FFN → residual stream → next-token prediction. LLMs essentially stack Transf…
A Chinese science tweet that intuitively explains the core chain of LLMs (Large Language Models): from token, embedding, position encoding, attention, FFN to residual stream and next-token prediction, helping readers without a math background understand AI papers.