The main question this raises for me is whether David is daft or whether he's lying. It's at least one or the other. Let's go through what David's said and why he's completely wrong and how, as a result, he's spreading misinformation. In the Rogan clip, David said "Sudden blindness. It's called NAION, which is now twice as more prevalent because of the GLP-1 drugs.... Sudden blindness has doubled seemingly because these meds [referring to GLP-1RAs]." To support this claim, David must provide data showing that the population rate of "sudden blindness" (and perhaps just NAION) has doubled. Does he? No. What he links, instead, is a study that compares small groups of people at one particular ophthalmology clinic who were taking GLP-1RAs to those taking either other diabetes medications or other obesity medications. He also linked a press release on a conference presentation to a similarly-weak analysis. Can you grasp the difference between what he should've linked and what he did link? He linked non-causal comparisons of groups that are very different from the general population, but he should've linked EHR or alternative monitoring-based prevalence estimates, like what I pulled. Now, if we take David's linked studies seriously, then the actual rate of NAION in society should be well beyond the actually diagnosed rate. The actual rate is 2-10 per 100,000. In one of the studies he linked, it's almost 2% of the diabetics on non-GLP-1RA medication and almost 9% on GLP-1RAs. Given the number of Americans who have used GLP-1RAs by now, you should know many people who went blind. But you do not, because these studies don't apply to the general population. I pointed this out before in a post linked in the thread that David is replying to, meaning that in addition to being daft, a liar, or both, he's also not circumspected, because his specific links were already shown to lead to absurd conclusions. In fact, the result was also shown to not hold up when the analysis is done in large, population-representative samples with better controls: My conclusions subsequently replicated again: And again: David's only holds up in weak studies with small samples, poor controls, and results that imply that HUGE portions of the population should've already gone blind -- and they have not, so they are obviously wrong. This is part of what I was getting at in the thread: we can sanity check these studies. If they imply that there should be an epidemic of blindness, we can look to see in descriptive data if there is such an epidemic. Since there is not, we know the studies are flawed in one or more ways that prevent generalizing from them. Get it? David didn't. Instead, he decided to go on one of the most popular podcasts and fail to even describe his referenced studies correctly. David's "sudden blindness" remark was actually fortuitous for me, because using all forms of acute blindness, we have enough power to detect a trend lately, if there's been one. There hasn't, as I noted. Looking at NAION specifically, we're left with a different conclusion: there's not enough power to detect a doubling (there is more than enough to detect what David's cited studies imply though!). The numbers have not drifted at all with age controlled (they're nonsignificantly down), which means that there's not even a reason to suspect the incidence is up. That also means there's not a reason for David to have said what he did. He just yapped without underlying data to support what he said, and data actually contradicting it. David's comment started ironically, by accusing me of having a financial conflict of interest. Ironically, he has a financial conflict of interest, in that he is developing a therapy that he believes will help with this issue that isn't actually an issue. So, he's financially incentivized to play it up, even though doing so is lying and is harmful because it gets patients to avoid good drugs on the basis of alleged side effects that would, at most, be extremely rare, and are more likely not present at all. So, is David dumb? Is he lying? Or is he a bad actor? Of course, he can be all of the above.

Under Biden and Obama, SpaceX got no reprieve. They had to do absurd nonsense all the time in order to do anything. Under Obama, the FAA made SpaceX write a book report on local activity in the Mexican-American War.

My god. The U.K., man... A journalist asked Arday about his fraud/plagiarism/issues, and Arday went to the police claiming that he was being harassed. So, citing Arday likely lying about the journalist's actions hurting his mental health, they told the journalist to back off.

After a supposedly 111-year-old man was found to have been dead since 81 and unreported so his daughter could keep collecting his pension checks, Japanese authorities began investigating the country's oldest old. They found an insane situation.

This is amazing. The White House waived Jones Act requirements in mid-March, and it's generated billions in economic activity and revealed tons of natural trade routes that were impossible while the Act was in effect. We really need to permanently remove this stupid law!

There's an asymmetry in how liberals and conservatives evaluate one another. Liberals dehumanize conservatives more than conservatives suspect. Conservatives dehumanize liberals less than liberals suspect.

The big news in biotech is that the EU just approved a cochlear implant for eyesight. Age-related macular degeneration affects ~200m people globally, and the "PRIMA" lens turns light into signals your surviving retinal cells send to the brain, restoring the ability to see again!

I have to assume this is a parody account given how often it tweets out extremely obvious lies. This is basically one of those gonzo news 'Biden was having lunch with lizardpeople!' accounts, but taking itself seriously.

While it would be nice for data centers to do all these things in every instance, this is a bafflingly anti-development standard. If this was applied rigorously to all infrastructure buildouts, then America would basically never build anything requiring power. Slopulism!

If you ask a frontier LLM a multi-hop reasoning question, e.g., "Who won the Nobel Prize for Chemistry in (1900 + Mozart's age when he died)?", it usually can't answer correctly immediately (no thinking) BUT if you ask the same question & append 300 dots, suddenly it can answer?

Every few weeks, someone on the TL rediscovers polyphasic sleep, rewrites their schedule around the idea, then ruins their whole life and alienates themselves from their family, friends, and job. SLEEP NORMALLY.