# Related Trends Back to [[Ribosome Pausing and Cotranslational Folding]] · Post: [[Post Analysis - thegreatest_sv on Ribosome Pausing (2026-10-23)]] · Evidence: [[Research and Evidence]] ## 1. "AI cracked open X" posts that re-label old science **Pattern:** take a real, often decades-old finding, attach the name of a famous AI lab, and present it as a new discovery. Here, Anthropic's genuine Sep 2026 structure-prediction work (which included a 70S ribosome *structure*) became "Anthropic cracked open the ribosome and found it stalls on purpose." Replies caught it within hours (@coolerwave, @Jim76638473, @danIsomLab, @grok). - **Tell-tale signs here:** no paper link, no named researchers, a four-step "method" in all caps, a profane hook, and a product call to action at the end. - Compare the same "big claim, no primary source" pattern in a different field: [[Expanding Earth and Plate Tectonics]]. ## 2. The AI-tools content economy on X - The author's week (Oct 18–24) is 19 original posts and 6 reposts of his own posts, mostly quoting his own long Articles. **7 of the 19 originals quote one Jev article**, each with a new hook ("Stanford AI researchers…", "Jev traders", "top 5 commands", this ribosome post). See [[Users/@thegreatest_sv/thegreatest_sv-x-profile|profile]]. - **Bookmark-heavy engagement:** this post drew 242 bookmarks against 270 likes; the Jev article drew 1,143 bookmarks against 471 likes on about 1.2M impressions. "Please bookmark, RT and follow" is written into the article. A high bookmark-to-like ratio is typical of "save this" listicle content ⚠️ (my interpretation, not a measured benchmark). - **Decorative "data" video:** an AI-styled dashboard with live-looking numbers and code adds credibility without content. The author declined to say what made it ("Secret"). - **Unclear sponsorship:** whether these Jev posts are paid is not disclosed either way ⚠️. ## 3. New AI model categories being name-dropped - **Jev / "System One" models** (TypeSafe, launched Sep 15, 2026): models that return typed decisions instead of text. Useful for routing and classification, which is why "Claude + JEV" (generator + router) is a coherent *product* pairing, but not a scientific instrument. See [[Research and Evidence#5. What Jev is]]. - **AI-accelerated structure prediction:** Anthropic's Claude-optimized biomolecular models and "Big" mode (Sep 2026) and its protein binder design result (Aug 2026). This is the real AI-biology news the post borrows from. ## 4. Codon optimization under scrutiny in therapeutics - Gene therapies, recombinant proteins and synthetic mRNAs often use **codon-optimized** sequences to raise expression. The concern that this erases native pauses and changes folding or immunogenicity is long-standing: [Mauro & Chappell 2014](https://doi.org/10.1016/j.molmed.2014.09.003); FDA CBER work on factor IX ([Alexaki et al. 2019](https://doi.org/10.1038/s41598-019-51984-2)); [Sauna & Kimchi-Sarfaty 2011](https://doi.org/10.1038/nrg3051). - Alternatives: **codon harmonization** ([Angov et al. 2008](https://doi.org/10.1371/journal.pone.0002189)) and design tools that preserve slow-translating regions. This is the real-world version of the post's "we may have been optimizing away part of the instructions." ## 5. Ribosome profiling plus machine learning - Ribo-seq ([Ingolia 2009](https://doi.org/10.1126/science.1168978)) and disome-seq ([Zhao 2021](https://doi.org/10.1186/s13059-020-02256-0)) produce codon-level maps of where ribosomes slow and collide. ML models trained on these data now predict elongation speed and its link to folding ([Aguilar Rangel, Stein & Frydman 2024](https://doi.org/10.1126/sciadv.ado0738)). - The field is still arguing about effect sizes: a 2026 preprint finds codon-supply metrics add little once amino-acid identity is controlled ([Hassan et al., bioRxiv](https://doi.org/10.64898/2026.09.27.753847), not peer reviewed). The post presents as settled something researchers still debate. ## 6. Ribosome collisions and quality control - The flip side of "pauses are good": long stalls lead to collisions, sensed by ZNF598 and others, then no-go decay and ribosome-associated quality control ([Doma & Parker 2006](https://doi.org/10.1038/nature04530); [Simms et al. 2017](https://doi.org/10.1016/j.molcel.2017.08.019); [Juszkiewicz et al. 2018](https://doi.org/10.1016/j.molcel.2018.08.037)). The post's "scientists treated stalls like a problem" refers, in effect, to this real research line. ## 7. Grok replies as a fact-check layer, with limits - @grok's English reply correctly said the biology is longstanding and that Anthropic did not discover it. Its Korean reply to @greenich_ repeated the post's claim that the video shows a "ribosome analysis flow," which it does not. Read automated replies as a lead, not a source.