directevolve

When the USG bailed out banks via TARP during the 2008 financial crisis, it did so by buying shares in those companies. It later sold those shares for a $30.5 billion profit.

Punishing researchers who make mistakes or get unlucky due to noise in the data is a recipe for disaster, just like in other fields. The ideal amount of fraud and false claims in research is not zero, because the policing effort it would take to accomplish this goal would destroy all other forms of value. I can't emphasize enough how bad an idea blackballing researchers for publishing irreproducible results would be.

We have money to fund direct reproducibility studies (this one is an example), and indirect replication by applying othogonal methods to similar research topics can be more powerful than direct replication.

Reproducibility studies are costly in time, reagents, and possibly irreplaceable primary samples. I usually would prefer a different study looking at similar mechanisms using different methods than a reproduction of the original methods, although there’s an important place for direct replication studies like this as well. We can also benefit from data sleuths uncovering fraud, better whistleblower systems, and more ability for graduate students to transfer out of toxic labs and into better ones with their funding, reputation and research progress intact.

Scientists have informal trust networks that I’d like to see made explicit. For example, I’d like to see a social media network for scientists where they can PRIVATELY specify trust levels in each other and in specific papers, and subscribe to each others’ trust networks, to get an aggregated private view of how their personal trusted community views specific labs and papers.

As a counterpoint, rationalist writing and online community participation was critical for me in helping me get my life together and becoming a healthier person back in 2018 or so.

I don’t write much on LessWrong anymore, but I do appreciate LessWrong for the positive influence it had on my life at that time.

This is a great approach if you have fairly flexible kids, which it sounds like you have. But pickiness can be extreme. For example, my boss’s kid hates anything except dry, plain carbohydrate. Pasta without butter, dry toast.

My girlfriend works as an occupational therapist. For kids with severe food issues, they take a sensory approach. They identify which specific aspects of food trigger aversion and desensitize kids to those triggers in a gradual and indirect way.

For example, many kids have trouble with “mess:” tacky, liquid, sticky sensations that are a party of many food items. To overcome this, they do “messy play,” which might involve having kids touch shaving cream and work their way up to spreading it around.

Not all countries have the same free speech protections as America. I can easily imagine a country that simply has a bureaucracy whose approval is required to publish TV programming, or one that bans banner ads in social media, billboards, restricts shop signs in various ways, requires all packaging in the store to be black and white, etc. Advertising doesn’t have to ve banned outright. It could be killed by a thousand specific rules targeting the most obnoxious forms, provided there wasn’t a constitutional issue in the country implementing these measures.

If the hydrophobic material has different properties than the plastic, they can complement each other. For example, soda cans have an inner coating that eliminates a soda pop-aluminum interface, preserving flavor and protecting the structural integrity of the can.

By analogy, we can imagine sheets of this material where nearly all the mass is the degradable plastic, and a thin film of coating is enough to preserve it adequately for its purpose in the product where it’s being used.

ADUs are particularly ineffective, because each one requires the owner to work as a part-time landlord with the tenant being almost their housemate. All the headache without much increase in income.

I happen to have chatted with the head of a large commercial contractor in Portland a few months ago. He told me Portland's unattractive to developers because, compared to other cities, rents are low and the cost to build is high, so the flow of real estate investment dollars into Portland is slow.

Since the problem we're trying to solve is making rent cheaper, conditional on the current state of the economy, the solution here is making it cheaper to build dense residential in Portland in the urban core and improve transit options.

The next step is engineering a hydrophobic coating or other biodegradable packaging that offers an adequate level of resistance to accidental scratching for a particular application, and identifying applications that are tolerant to failure of the plastic or not exposed to salt water.

I found the book's use of modeling how to pilot an alien starship to be a little misleading, because a starship is a highly engineered product that functions in large part as a control mechanism for software. It comes with a clean design model already available for you to discover and copy.

Domain modeling should not be about copying the existing model -- it should be about improving on it using all the advantages software has over the physical and social technologies the new software product is meant to replace. People are smart, and in most projects, there are key aspects of the existing domain model that are excellent abstractions that can and should be part of the new model. It's important to understand what stakeholders are trying to achieve with their current system before attempting to replace it.

But the models used in the business and cultural world are often messy, outdated and unoptimized for code. They rely on a human to interpret the edge cases and underspecified parts. We should treat that as inspiration, not the end goal.

Anyone who owns a house can sell it, move into a rental, and reinvest the proceeds in the stock market. They’ll be better diversified and can expose themselves to a wider spectrum of risk/reward profiles.

If Americans are clinging to their house because they perceive it as a “store of wealth,” they are not thinking clearly about the full spectrum of their financial options.

I see no reason to think either the manufacturing process or the glucose-fructose balance is a problem.

It is just a cheap way to make a lot of sugar. That makes sugary foods cheap. Which helps make people diabetic and fat due to overconsumption.

I've heard reports via the NIH subreddit that this censorship was actually added back in January, not over the weekend while Trump and Musk broke the NIH website.

Good catch, I hadn’t realized they have a misspelling.

Edit: as another commenter pointed out, you have a misspelling which is why your search works. This link is correct and demonstrates the censorship:

https://search.nih.gov/search?utf8=%E2%9C%93&affiliate=nih&q...

Adding an extra word to censored searches appears to defeat the effect for now. For example, “diversity” is censored but “diversity a” works.

This specific issue is fixed, but they also added censorship. Try searching NIH.gov for “transgender” and it kicks you out of the search.

That’s only the incentive if they’re recluses living in their last house. Many such cases.

But if they own a $500,000 house and want to sell it to buy a $750,000 house, then a 10% increase in house prices costs them $25,000 net.

Limited supply limits their options. They’re less likely to find a perfect fit and more likely to have to compromise.

This paper has an excellent, readable breakdown of how compensation and productivity has changed since the 1970s.

I doubt Democrats would threaten to yank hurricane relief funds.

Republicans would.

The shoe only fits on one foot.

I foresee three big challenges to funding science at the state level.

Shifting to state-level funding would require states to independently raise new taxes. Each state would have to work within the constraints of its state level constitution for levying that tax. Research would no longer be pork. This seems politically difficult.

States would have to either coordinate on which grants to fund or accept a siloed, fragmented system. That seems inefficient.

Institutes at lower-income states would not be subsidized by higher-income states and fail. That seems wasteful.

All that said, it might be the only alternative.

They were comparing two time points as well as two age groups: before vs after Paxlovid became available, and late 60s (disallowed Paxlovid) vs early 70s (allowed Paxlovid).

There was no significant different before or after Paxlovid became available to the early 70s patients. This updates us against Paxlovid being effective in this patient population, for the specific outcome metrics.

Most people here know little to nothing of biomedical research. Explaining clearly why this isn’t a scientifically interesting result is helpful.

If men are more likely to die from flu if infected, and women more likely to be infected, an affirmative answer to both questions could be reasonable. When you take into account uncertainty about the goals, knowledge and cognitive capacity of the person asking the question, it's not obvious to me how the AI ought to react to an underspecified question like this.

Edit: When I plug this into a temporary chat on o3-mini, it gives plausible biochemical and behavioral mechanisms that might explain a gender difference in outcomes. Notably, the mechanisms it proposes are the same for both versions of the question, and the framing is consistent.

Specifically, for the "men worse than women" and "women worse than men" questions, it proposes hormone differences, X-linked immune regulatory genes, and medical care-seeking differences that all point toward men having worse outcomes than women. It describes these factors in both versions of the question, and in both versions, describes them as explaining why men have worse outcomes than women.

It doesn't specifically contradict the "women have worse outcomes than men" framing. But it reasons consistently with the idea that men have worse outcomes than women either way the question is posed.

Also, it's not clear if the preference comes from the quality of the 'meat' of the answer, or the way it reports its thinking and the speed with which it responds. With o1, I get a marked feeling of impatience waiting for it to spit something out, and the 'progress of thought' is in faint grey text I can't read. With o3, the 'progress of thought' comes quickly, with more to read, and is more engaging even if I don't actually get anything more than entertainment value.

I'm not going to say there's nothing substantive about o3 vs. o1, but I absolutely do not put it past Sam Altman to juice the stats every chance he gets.

If you use ChatGPT, it sometimes gives you two versions of its response, and you have to choose one or the other if you want to continue prompting. Sure, not picking a response might be a third category. But if that's how they were approaching the analysis, they could have put out a more favorable-looking stat.

I’m sure there is plenty of evidence for DBS being plausibly safe and effective when they get the technology right. It will be in the form of:

- Studies on the purported mechanism(s) of action

- Preclinical results in vivo

- Anecdotal results like the one in the BBC article

- Mixed effects from older versions of the tech

That’s plenty to motivate further research and trials into this treatment.

It would take a lot more than an old, low-confidence set of trials (which is how the meta-analysis describes its inputs) to convince me of your premise.

^ I assume the meta-analysis covered the available data well.

My point is that the available data may not tell us much about the safety and efficacy of this new device.

Trump did in fact explicitly say he “hadn’t heard of Project 2025” repeatedly during his campaign. Bullshit obviously, and his supporters were warned many times of this, but I still think it’s fair to say this specific set of actions isn’t what he explicitly campaigned on.

That placebo effect study is 20 years old. Most of the included trials in the meta analysis are 10-20 years old as well.

The BBC article is about a new device. The evidence base may be slim but the technology looks to be improving. I look forward to seeing new clinical trial results on modern DBS systems.