@FinanceYF5: 2/ He never looks at benchmark numbers when evaluating models. The only thing he truly cares about is: [The shape of the model's thinking] — How deeply can it understand user intent? — How far can it iterate in its thinking? — Does it make you feel like there's someone on the other side? Fable gave him this sense of aliveness. 'It feels like returning to 2023'
Summary
This tweet emphasizes that when evaluating AI models, one should not only look at benchmark numbers but focus on the model's 'shape of thinking' — the depth of understanding user intent, the ability to iterate in thinking, and the feeling of 'someone on the other side'. The author believes Fable excels in this regard, reminiscent of the experience in 2023.
View Cached Full Text
Cached at: 06/22/26, 09:40 PM
2/🧠 He never evaluates models by looking at benchmark numbers
The only thing he truly cares about is:
【The shape of the model’s thinking】 ——How deeply it can understand user intent ——To what extent it can iterate its thinking ——Whether it makes you feel like “there’s someone on the other side”
Fable gave him that sense of liveliness “Like going back to 2023” https://t.co/Us81ixnO1F
1/ Has the window for entering the AI capability race already closed?
There’s only a three-year window to join this competition: from February 2023 to February 2026. That window is now shut.
Most countries have permanently missed the boat.
This isn’t pessimism It’s the conclusion someone reached after using Fable
Why did one model make him feel this way?
3/ He believes the leap in AI capabilities over the past five months
Comes not just from tool improvements like Claude Code
But because of 【Mythos】—a new model from Anthropic That completed training this February and quietly changed the entire R&D rhythm
Key judgment: The leading models are helping to train the next generation of leading models
This cycle has already started
3/ Some rush in, others busy building walls
Musk approached the frontier in 26 months, one of the few who seriously tried.
Europe spent a lot of time setting up regulatory barriers, and by the time they woke up, it was too late.
4/ Your “national AI” might just be a shell
Many countries’ “homegrown models” are just a layer wrapped around models from US or Chinese labs.
Once supply is cut off, the effect is like an air raid—instant paralysis.
5/ Computing power will be regulated like uranium
Advanced chips will be licensed, monitored, and kept within the US; open source can help, but it can’t save countries lacking computing power and talent.
As open source approaches the frontier, global regulatory crackdowns will follow.
6/ His final conclusion, and the most jarring sentence:
“If you are not yet building it You will only be an onlooker”
AI is not about writing code, copyright, or art It’s about the reshaping of civilization, society, and humanity itself
This is the correction he made to his timeline After using Fable for one day
Read the original:
That’s all
If you like this topic:
- Follow me (@FinanceYF5)
- Like + retweet the first post below
Similar Articles
@FinanceYF5: 2/ First change: Give only the goal, not the steps. With previous models, you had to spell out exactly how to do it, or they would go off track. Fable 5 is the opposite — the more freedom you give it, the better it performs. Every step you dictate to it is actually overriding its better judgment with your own.
Discussing tips for using the AI model Fable 5, advocating for giving only the goal without steps, and granting the model more freedom to achieve better results.
@FinanceYF5: It is said that a 'certain new model' from zAI is at least as strong as Fable 5 in cybersecurity capabilities. Chubby did some research but only found a Wall Street Journal article. However, that article did not mention a completely new model; instead, it discussed GLM 5.2 as a relatively recent model…
According to rumors, a certain new model from zAI is at least as strong as Fable 5 in cybersecurity capabilities, but information is limited, only citing a Wall Street Journal article discussing GLM 5.2.
@FinanceYF5: Anthropic's outlook appears bleak: Fable 5 is losing market share to more affordable open-source models. According to the Financial Times, in Ramp's enterprise AI spending data, Anthropic's top model Fable 5 only accounts for 11%. The reason is simple: Fa…
According to the Financial Times, Anthropic's strongest model Fable 5 captures only 11% of enterprise AI spending, being overtaken by cheaper open-source models due to its high price.
@CMGS1988: https://x.com/CMGS1988/status/2074488576356876585
The author shares personal insights on the relationship between AI and infrastructure, arguing that while AI amplifies execution capabilities, it cannot replace human judgment on complex trade-offs and organizational issues. The author illustrates this with their own experience developing the cocoon/sandbox project using models like Fable 5.
@mylifcc: https://x.com/mylifcc/status/2073053339714212161
The article emphasizes that when using strong reasoning models like Fable 5, one should prioritize auditing and reconstructing one's personal work operating system (such as coding, AI lab, content synthesis, etc.) rather than directly using them for coding. Through system-level upgrades, a compounding effect can be achieved, significantly improving the quality and efficiency of all subsequent outputs.