Get the latest tech news
Differences in link hallucination and source comprehension across different LLM
I've been waiting for someone else to talk about this, but I guess it will be me
Major reviews of the entire field conclude that the overall quality of evidence for long-term ADHD medication benefits is "very low," meaning the MTA's findings are part of a broader landscape of scientific uncertainty ( The Lancet Psychiatry). Other longitudinal and registry studies (e.g., Swedish criminal-registry work) find domain-specific benefits such as lower conviction rates during medicated periods, suggesting the MTA’s outcome set did not capture every potential long-term effect ( NEJM). You could run this automatically, hide the intermediate output, just show this result and have a better summary of this than just about any press coverage out there (if you’re a medical researcher looking at this and thinking, oh but I would have written it this way, go ahead and point me to the public-facing reporting that explains this better — I’ll wait).
Or read this on Hacker News
