Many users criticize AI models for failing to capture writing style and understanding in tests because outputs lack nuance, depth, and original thought despite mimicking structure.
Based on 14 visible X reactions from 27 accounts; directional sample.
Ask a question below.
Published answers will appear here.
Fired our content team in March. Replaced them with GPT-5.6. The output was technically correct - right word count, right tone score, 91% readability. My VP of Marketing said it 'read like someone summarizing a dream they half-remember.' We published 340 articles last quarter. Organic traffic is down 62%.
@GergelyOrosz Tested the new models again: fed them my old interviews and asked to write an article in my exact style. Fable & GPT-6? Epic fail 😂 Tons of fancy words, zero soul or real understanding. AI still can’t copy real voice yet.
@GergelyOrosz Related: I’ve interviewed swe’s who are using cluely or some variant and it’s terrible just how terrible LLMs are at interviewing today. There’s something human that these models / harnesses just struggle to capture.
@GergelyOrosz AI writing scans nicely, but falls apart as soon as you start to try to understand the points its making It's hilarious how bad such a "smart" thing can me at making coherent points
Fired our content team in March. Replaced them with GPT-5.6. The output was technically correct - right word count, right tone score, 91% readability. My VP of Marketing said it 'read like someone summarizing a dream they half-remember.' We published 340 articles last quarter. Organic traffic is down 62%.
@GergelyOrosz Tested the new models again: fed them my old interviews and asked to write an article in my exact style. Fable & GPT-6? Epic fail 😂 Tons of fancy words, zero soul or real understanding. AI still can’t copy real voice yet.
@GergelyOrosz Related: I’ve interviewed swe’s who are using cluely or some variant and it’s terrible just how terrible LLMs are at interviewing today. There’s something human that these models / harnesses just struggle to capture.
@GergelyOrosz AI writing scans nicely, but falls apart as soon as you start to try to understand the points its making It's hilarious how bad such a "smart" thing can me at making coherent points
I like this take! And I am also amused at how bad these models are at writing (that is worth reading) but then very good at coding! https://x.com/ScottWRobinson/status/2076758105057014270
With every new model, I put it to the test in writing - to see if I could do what I do - I give it a bunch of interviews I did on a topic, and ask to write an article in my style. The models fail spectacularly, Fable and GPT-6 no different. Lots of words, but no understanding
Many users criticize AI models for failing to capture writing style and understanding in tests because outputs lack nuance, depth, and original thought despite mimicking structure.
Based on 14 visible X reactions from 27 accounts; directional sample.
Ask a question below.
Published answers will appear here.
With every new model, I put it to the test in writing - to see if I could do what I do - I give it a bunch of interviews I did on a topic, and ask to write an article in my style. The models fail spectacularly, Fable and GPT-6 no different. Lots of words, but no understanding