I'm open to this correlation. But I wonder about causality. Adverse childhood experiences are significantly associated with both impulsivity and future earning potential. Breaking the cycle of poverty is difficult, but I'm not an expert on where the best investments to solve this problem would be.
> But I wonder about causality. Adverse childhood experiences are significantly associated with both impulsivity and future earning potential.
There's fairly good statistics that a large portion of the people living in a cycle of poverty and drug addiction in Vancouver's downtown eastside suffer from a number of documented risk factors:
a) fetal alcohol syndrome
b) documented abuse by parents at an early age, or became wards of the state and entered foster care system, were not placed with any family members
c) drug/alcohol abuse by parents
d) molested by parents, relatives, "friends" of parents at an early age
If the parents have poor impulse control then the children often do as well. This is a selection effect; children growing up in poverty often inherit a significant contributor to their parents' persistent poverty.
People who are incidentally in poverty usually have children that aren't. I grew up in pretty abject poverty and saw this pattern over and over again too. My parents didn't have poor impulse control, it was circumstantial. All of my siblings were able to grind their way to a successful life despite having the deck stacked against them -- and none of them work in tech.
It is extremely frustrating. I have persistently poor friends who are unambiguously above average in both intelligence and work ethic, and with great personalities. Yet despite that, they constantly sabotage their future with poor impulse control. With good impulse control they'd be solidly middle class at a minimum.
I don't know how to fix the above situation. I've tried many times. They can't really explain why they make the stupid decisions they do, it just kind of happens.
I also wonder if its rational behaviour at the end of the day. If you don't have anything to your name, why delay gratification? Its not like you'll get benefits of delaying, and you may very well suffer a loss if you wait. I think you have to have a certain amount of wealth before the types of behaviours that actually make you wealthy make sense in the average case.
The original marshmallow test fails for this reason, but also it's true that poor impulse control makes it harder to break out of situations that do not reward delayed gratification.
> Adverse childhood experiences are significantly associated with both impulsivity and future earning potential.
But that’s just another correlation without a proof of causation.
There’s no clear definition of what causation is. You can invoke the tired “correlation doesn’t imply causation” any time you want, but the same quip disqualifies any alternative causation mechanism you propose, unless you serve up an actual criterion for distinguishing the two along with it. Asserting that one seems more plausible than the other isn’t enough.
Pleased to see Roget's thesaurus here. It's a fantastic resource without a great digital equivalent. I've found LLMs helpful in a pinch (and it's certainly better than a standard thesaurus), but Roget's is still the book I turn to for tricky word-picking problems.
I've never owned a copy of Roget's, and now I'm curious/tempted to get one, but my go-to digital word picking assistant has always been OneLook Thesaurus: https://www.onelook.com/thesaurus/
Hey, I'm the developer of OneLook. I'm so glad to see it here!
The concept clusters on the site are directly inspired by Roget's work. Like the author, I see a thesaurus less as a synonym dictionary and more as what the word literally means, a storehouse of treasure. I'm in the midst of a long-overdue data refresh across OneLook right now, so I'd be curious to hear your feedback once it's done.
Very cool. This will likely become my default digital source. Much more extensive than the other online options. I still prefer the discrete clustering of Roget's, but this blows away anything else I've found.
Twenty years ago, OneLook Reverse was how I took college papers from a B to an A through intimidating the TAs with absurdly specific yet exactly fitting three dollar words. That they thought I knew the field led to great friendships (and jobs) with some of them. I owe you big time and thank you!
Note that this version of Thesaurus-1911 has been supplemented with over
1,000 words not present in the original 1911 edition, but many modern words
are still missing. About 1500 verbs (out of 6500) which can be found in an
80,000-word spell-checker are absent from this work. The deficiency of nouns
is probably much worse, especially on technical topics. Of 40,000 unique
words contained in the original text, 12,000 are not recognized by a
spell-checker. Most of these are foreign words (primarily Latin), and many
are obsolete.
I never knew Roget's was organized like that and I have had several different editions over the years. I'm one of those people who always used it as a compendium of synonyms.
A lot of books that call themselves Roget’s have abandoned the dual structure (classification-index) in favor of a single index structure built like a dictionary, basically constructed for people who want a different word than what they would normally use, as mjd writes, rather than people who are interested in the store-house (thesaurus) of the English language.
I was looking for it after reading the post, but there seem to be so many different versions and edtions. Harper, Penguin. College. The New. The Concise. The International. Some say they’re organized alphabetically. How do I know which one is the kind that the article talks about, if I buy used online and cant look into it?
The one in TFA is Harper and Row, and the later 8th Edition, published by Harper Collins, is also organized the correct way. I think you'll be safe if you can get a Harper version.
That's the one I have. I've had an older version in the past but misplaced it years ago. This one does the trick, but like the article author says: it is a weighty book. (I like that.)
Great code needs great understanding and agents need excellent guidance. Even in my current solo-dev work, I can't imagine making a production commit I haven't read until I understand it. I own the consequences of my code; that's a responsibility AI agents can't take.
Agreed with Reading! This alone isn't enough though. PreLLM too it wasn't just reading code to review. Someone did the hard work or crafting the code and each unit test would tell you the weird corner cases to deal with and factor that into changing your code. One person owned a part of the codebase and was an expert.
Not to mention reading isnt easy when the velocity of code pumping in is 3-5x more. Its exhausting and reading becomes skimming.
The expert is now outsourced to LLM. If someone asks me about a bug in a system I made N years ago, I usually have a hunch what the problem might be, now it feels I'm lost in my own codebase (even if I really read the code). Similar to why math books have exercises and not only explanations.
Guess it's not an issue as long as you have access to the models and someone who likes prompting.
So true. Writing the code was so helpful for learning what it meant. It takes much more investment to go back to the code LLMs write to figure out what's actually happening and weigh everything.
Yeah, but in the corporate world now, you're being judged by how fast you get shit checked in. So, if you need to understand every line these days... you're going to end up in the bottom 25% of the company pretty quickly. Not saying you're wrong, just saying how fucked things are becoming.
Perverse incentives have always been the purview of bad management. Measuring things that don't cause success is easier than figuring out what actually matters. (It's also why I don't practice corporate medicine anymore.)
Conversely I'm a solo-dev and I ship a lot of slop I don't even look at. Granted my work is just basic CRUD apps, and I focus my efforts on validating important consequences (like does this break accounting invariants or something)
I like the optimism in this, and the rightful identification of many important bottlenecks.
It's important to recognize that many of the remaining medical problems are very hard to solve. And while I'm still hopeful for relatively simple answers to some, successful interventions for others often require changes across multiple systems, and at economic or social costs we haven't been willing to bear. Identifying new technical solutions across all stages of research is important. But we also struggle to implement many basic interventions that already work.
These type of articles always start with low hanging fruit and then ignore or overlook that some problems are not at all like that. Different mechanisms altogether, like cancer or Alzheimer's. These are way harder.
I like this form of parenting. Make your life better for your kids, but reintroduce some challenge so that they learn to appreciate the comforts of life.
Software that works year-after-year has never been a commodity. It's boring on the surface. It doesn't get the flashy posts. But I'll choose reliable over new in almost all cases.
This is not at all what the article says. The argument is that the current classifications don't work, and that consensual sex should never be considered paraphilic.
This funding campaign is questionable. Investors should have some access to profit and loss information before investing. I don't have a major problem with unaccredited investors jumping in and losing money, but I do think that they deserve at least minimal transparency.
I'm not far above the accredited investor line and have a few AngelList investments. They've always provided cash flow, burn rate, projected revenue goals, and how they plan to hit them. I know I can't perform due diligence and know I'll probably lose the 1-2% of my portfolio in angel investments, but I would run from any company that doesn't offer real information.
I'm a big fan of Substack and hope that I'm wrong. I just hope that trying to reach venture expectations doesn't tank what could be a fantastically lucrative SMB.
It's provocative in rats, but there's lessons from similar attempts that need to be considered. Cocaine-specific vaccines have been in development for years, and they haven't been successful. The biggest problem is that you need people to maintain an enormous concentration of antibodies against the drug so that even minute quantities are captured within seconds of entering the blood stream. Even if you can achieve an initial therapeutic levels, you need to constantly boost the vaccine to maintain high antibody titer. All of these attempts have targeted people with cocaine use disorder and are able to focus on shots every few weeks to months. We're nowhere close to developing something that would work as a routine preventative vaccination.
We do have a once a month injection of naltrexone that can block all opioids, which can be effective in the right person. A targeted approach to fentanyl alone would probably work in people who don't have an opioid use disorder (and therefore wouldn't be as prone to substitution with an alternative opioid). This is an interesting step, but our understanding of immunology is far too limited to make this realistic in the near term.
