X user roon (@tszzl) drew over 160,000 views on May 24 after noting that Claude 3.5 Pro’s growing adoption among professionals — illustrated by introducing an “elite lawyer friend” to the model — stems less from benchmark superiority and more from Anthropic’s deliberate strategy of packaging the model for specific industry use cases. Roon wrote that “the models do not sell themselves” and credited Claude’s vertical marketing as a key differentiator. Replies in the thread noted that legal workflows still surface hallucinations and require iterative prompt adjustment before the model becomes reliably useful.
Sentiment analysis of the 155-comment thread skewed roughly 74% negative, with critics arguing that vertical packaging masks persistent reliability gaps and that the model still fails on complex legal reasoning. A minority of respondents praised Claude’s performance in document-heavy legal tasks. The exchange reflects a wider tension in enterprise AI adoption: whether go-to-market execution can compensate for model limitations, and how much friction professionals are willing to absorb before AI tools deliver consistent value.