jcoq

Actually, I think my fundamental challenge with AI is that it has no common sense. The way it builds things, writes, and operates is out of touch with reality.

Incidents like hugging face are partly rooted in the lack of common sense. It still functions like a supercharged toddler.

I'd love to overcome this because it'd mean I spend less time guiding the the LLM to produce usable outputs.

Much of this is foisted upon us by visual designers who wandered into product design. It's a category error the profession has never quite corrected. (maybe more controversially, it's caused by having anyone with the word "designer" in their title on a project that doesn't need such a person - this category is larger than anyone thinks)

CBS mischaracterized this memo, which says basically that there will be fewer of some jobs and more of other.

It's a pretty bland memo and a thinly veiled advertisement for AWS.

Sort of an entertaining read. The cornerstone of the argument starts on page 55 - three purported statistical anomalies that appear to be well-known and debunked.

* One person sees 1 of 4 images and another far away guesses the image. Supposedly it's accurate ~33% of the time instead of 25%. Fun but it can't be replicated by anyone else. Selection bias, poor design, and bad meta analysis explain the difference.

* ~54% of the, a person can detect if they're being stared at, whereas we'd expect 50%. Again, cannot be replicated in better designed studies and bad study design and selection bias explain the other results.

* By a small but supposedly statistically improbably amount, people can get a dice to roll by "wishing" for it. Again, cannot be replicated and can be explained with p-hacking anyhow.

You could still still see a greater left-side incidence if arm melanomas were higher on the left arm even though trunk/hip thigh melanomas were evenly distributed.

The blog fails to answer the most basic question: do right-side sleepers with a metal bedframe aligned with a radio tower have a higher incidence of left-side cancers that can explain the general population discrepancy?

Agreed. I'm an extremely fit runner doing 50+ miles a week and almost all of my runs are in a super easy zone that is equivalent to an uphill walk for most people. A lot of this has to due with well developed biomechanics, but also superior vo2max.

Most health focused runners should be walking at an incline or running downhill except for 1-2 high intensity interval days. There are huge benefits to low intensity activity and higher doses are almost always better. The sort of intense, grueling runs that most people do are likely counterproductive. It's like running a race every day as a workout. It's just not sustainable and unhealthy at high doses.

(Of course, if the goal is to be a runner and not simply be healthy, you'll need to push your body to a point where running is like walking but it takes a long time to do correctly.)

> connected very intimately

Do you mean hyperlinks?

They're not hiding anything: https://www.consumerreports.org/cro/about-us/our-partners/co...

If we give substance X to 17 people and 100% of them die, I would recommend against taking substance X.

Of course, it's not as black and white here, but sample size and effect size should be considered in relation to each other. Previous research and theoretical expectations should also play a leading role.

Apologies for misrepresenting your sources.

I think there are salary differences across careers for a variety of complex market reasons. If I had understood the degree to which money would dictate the security of my family, I'd have spent my years as a lowly paid graduate student and post-doc becoming a surgeon and not a mathematician. My bay area salary is much better than my tenure track salary but still a fraction of what I could actually be making. Nurses, too, are free to retrain.

I will concede that this issue is not personally important to me.

Nursing only requires a 2 year degree in the US and hardly any ongoing training apart from paid training provided by the hospital. There might be some additional certifications for certain fields, but that's it.

Most nurses are paid hourly, so the "long hours" is a moot point.

My nurse friends all make around 100,000 per year after overtime, which they're eager to get. The average full time salary without overtime is ~80,000.

For a 2 year degree, this is very decent money in most states and I think we'd struggle to name a career with such low education requirements and high salary.

I really appreciate the work of medical professionals and don't want to belittle their work. But I think nursing does standout as a counterexample to the point you're making.

Neither of your links mention nursing.

I don't think the mortality rate needs to correspond to the incidence of long term effects.

I was saying that I think it's better to assume you're not dishonest, just intellectually lazy.

> "The effect of vaccination was not systematically assessed."

I think you're either paraphrasing deceptively, or you are being intellectually lazy. Perhaps it is better to not assume malice...

The rest of the quote is:

"In total, 144 participants received an mRNA vaccination between the baseline and the follow-up scan. We performed separate analyses for participants with vaccination as well as for participants without vaccination. The results were not different from the findings of the full cohort as presented. The cardiac effects of vaccination require further research."

This is a remarkably stupid comment but not everything is "unreasonably effective". Mathematics was noted as unreasonably effective for modeling the universe because most areas of mathematics were not invented for the applications they meet... like discovering that your coffee maker doubles as an exceptionally good waffle maker.

Makefiles on the other hand, are not unreasonably effective in this sense. Makefiles are, in fact, _reasonably_ effective... like a coffee maker that brews coffee well.

I came here to say the same exact thing! These courses are excellent.

As a follow-up to those courses, there are also some courses about the middle ages by Philip Daileader that pick-up the story at the decline of the Western empire.

(Also, I'm deriving great value from the Wondrium app).

I think you're largely missing the point of this article, which mostly summarizes a novel technique for comparing gut biome diversity, and which goes through a small amount of effort to summarize the growing (if not clearly established) scientific consensus about the importance of healthy gut biomes.

I don't think it takes very much imagination to understand why the loss of diversity in an ecosystem might have negative implications for that ecosystem.

For whatever it's worth to you, Science magazine is a companion to the extremely prestigious Science Journal and it often tries to summarize the research in the journal. The tiniest amount of searching shows that the biologist whose work is being summarized a is one of the leading experts in this area.

This viewpoint manages to miss the purpose of nearly all software engineering.

We've already solved the tough algorithmic challenges relevant to most industries and those software engineers spend their time writing specifications that just happen to be programs. Their job is to have hours of discussion and think through edge cases to provide an exact description of the business requirements. Good engineering means thinking through edge cases and gray areas while the PMs and designers capture broad requirements from users and focus on high level details. You're not going to replace engineers in this capacity without human level, general AI.

Further, you cannot capture the needs of a tech-enabled business with natural language descriptions while remaining unambiguous and changeable. If you had an AI to write your programs, you'd still have engineers writing specifications into the AI (and even more importantly, deciding on those specifications down to the most intricate if details).

We will in fact see an explosion, or at least mild expansion, of software engineering into areas that are algorithmically challenging. AI will manage to solve the tough algorithmic parts and engineers will drive the AI through increasingly specialized languages. In this world, AIs will act like compilers.

But it's just super futuristic and naive to think that a PM will tell GPT-3 what to make like a Google Home and have it pop out the other side.

There have been other studies with similar results but the main drawback of these population analyses is that glucosamine use is more likely in people with healthy lifestyles.

Other supplements have shown benefits in these sorts of studies without standing up to more exacting experiment design.

So potentially interesting results but nothing worth getting invested in.

"What does it mean to get high? In a sense, it isn’t so complicated, when we do drugs, our “elevation” rises."

Clearly, drugs aren't doing anything for this author's ability to discern intelligent thought from vapid, intellectual posturing.

Climate change has turned into a sort of quasi-religious moral issue that blends with other issues of our day.

The thinking goes that, if only we could become pure and stop partaking in the evils of consumer capitalism, we might appease a hidden power and be saved from a myriad of bogeyman such as climate change.

This mindset fails to reasonably consider the certainty and enormity of the threat. Organized civilization is likely to end. Billions will die and we might very well become extinct.

The problem must be attacked with the full force of human intellect. It's so damn obvious that "wait for everyone to become super duper conscientious" is a fool's plan.

That argument is totally unconvincing... like saying not to treat some lung cancer because the patient will just smoke more.

Nearly all of the disastrous things we've done to the planet were understood to have negative environmental consequences and yet we did them anyhow. Meanwhile, a great deal of our environmental interventions have been incredibly successful.

So I'm totally unconvinced by this often recited and rarely supported mantra.

70% of the water in Arizona goes to farming. Our state ships hay to the Middle East. Subsidized water is a ridiculous subversion of market economics.

Blaming this on suburban households is like pretending that climate change would be solved if only consumers would be mindful of their carbon footprints.

Right? There's nothing surprising about getting scammed when you give out the 2FA code.

It wouldn't be very difficult to find a senior lecturer of Christian theology in any American university with borderline insane beliefs.

I wonder how much of this is lazy journalism coupled with the fact that non-believers might not want to speak out.

These all sound like pros... A project that makes you intimately familiar with a huge breadth of deep, foundational domains and that forces you to make opinions about esoteric concepts, all while pushing you to go deeper.

I would be terrified to see how some of the people I've worked with would abuse descriptors... thankfully they never knew about them(or it wasn't a python shop).

I've never understood the rationale that some developers employ for using `@property` (for purely stylistic) reasons and nobody has ever managed to give a very clear explanation of their criteria. Of course the concept is handy here and there, but some people make classes where every method is a property.

Long covid is definitely a scary prospect, but I think it's interesting that people also complain about long-term flu symptoms.

https://www.bbc.com/news/health-58726775

You can add multiple keys to the account.

That has almost nothing to do with it.

It's about convenience vs security, and specifically, auto-filled passwords across devices and applications. Anything that provides such a feature will be inherently less secure than an encrypted file.

If LastPass and others are to blame for anything, it would possibly be their marketing around the this tradeoff, though I think they avoid direct misrepresentation by just avoiding the issue completely.