You just have to double check the results whenever you tell Claude to extract lists and data.
99.9% of it will be correct, but sometimes 1 or 2 records are off. This kind of error is especially hard to notice because you're so impressed that Claude managed to do the extract task at all -- plus the results look wholly plausible upon eyeballing -- that you wouldn't expect anything to be wrong at all.
But LLMs can get things ever slightly wrong when it comes to long lists/tables. I've been bitten by this before.
Trust but verify.
(edit: if the answer is "machine verifiable", one approach to ask an LLM to write a Python validator which it can execute internally. ChatGPT can execute code. I believe Sonnet 3.5 can too, but I haven't tried.)
Some of them identify themselves by user agent but don't respect robots.txt, so you have to set up your server to 403 their requests to keep them out. If they start obfuscating their user agents then there won't be an easy solution besides deferring to a platform like CloudFlare which offers to play that cat and mouse game on your behalf though.
The (economic) purpose of Girls in Tech is to create more workers in software so that prices for labor will go down.
Given the layoffs of 2022-ongoing, labor costs in tech are dropping enough that interested parties aren't incentivized to increase the supply of labor further.
They don't do anything without $40 billion of air assets circling their position providing cover and get pissy when you tell them they have to shave and can't paint their weapons arbitrary colors.
Also they're all on drugs and will kill you if you report them for embezzling unit funds.
Posts like these are like the main character threads on twitter where someone says, "men don't do x" or "women aren't like y." It just feels like people outside of you who have no understanding of your context seem intent on making up rules for how you should code things.
Perhaps it would help to translate this into something more like, "what pitfalls do you run into if you parse `ls`" but it's hard to get past the initial language.
> A city slicker shoots a duck out in the country. As he's retrieving it, a farmer walks up and stops him, claiming that since the duck is on his farm, it technically belongs to him. After minutes of arguing, the farmer proposes they settle the matter "country style."
> "What's country style?" asks the city boy.
> "Out here in the country," the farmer says: "when two fellers have a dispute, one feller kicks the other one in the balls as hard as he can. Then that feller, why, he kicks the first one as hard as he can. And so forth. Last man standin' wins the dispute."
> Warily the city boy agrees and prepares himself. The farmer hauls off and kicks him in the groin with all his might. The city boy falls to the ground in the most intense pain he's ever felt, crying like a baby and rolling around on the ground. Finally he staggers to his feet and says: "All right, n-now it's–it's m-my turn."
> The farmer grins: "Forget it, you win. Keep the duck."
James Maynard appears regularly on Numberphile so if you'd like to hear some accessible mathematics from one of the authors of this paper I suggest you check it out:
As long as your python program checks for __name__ like this:
if __name__ == "__main__":
# Your code here
Then it should mostly "just work". Use the argparse module to parse command line arguments if necessary. Create a setup.py and use pip to install it, if you want.
There's a good reason for this. Computer numbers are not the same as numbers in mathematics. Floating point numbers are not real numbers and ints are not natural numbers.
"Programming" as a career is in an interesting position right now. Due to demand and the job itself, you can make a lot of money, work remotely, in a medium-low stress job, without interacting with the public, solving puzzles. That's the dream right? It really doesn't get any better than that for most people. The problem is that there are two types of people: those that enjoy coding, and those that don't. The people that enjoy coding are living the dream. I know I am. I mean I have it fucking good. All the benefits above, to the max.
The people that don't enjoy coding can still force themselves to do it... But they will find it lack luster. They are doing something they don't like, aren't too good at, for less money, at a job that took forever to find, etc etc. They won't have the same experience.
There are enough people that enjoy coding to tell everyone else "yeah this fucking rocks", and that certainly draws people in, but a lot of people will be disappointed with it.
Note that this requires giving the plugin permissions to read and write to all pages you visit. Browser extensions are routinely compromised by bad actors to steal banking passwords or bitcoin wallet keys.
I use tampermonkey, but I have it restricted to just one domain, which it complains about every time chrome is restarted.
People still don't realize how many things from the last fifteen years were LIRPs. Low interest rate phenomena doomed the moment investors can get a halfway-decent yield from fixed-income investments.
Where agents will potentially become extremely useful/dystopian is when they just silently watch your entire screen at all times. Isolated, encrypted and local preferably.
Imagine it just watching you coding for months, planning stuff, researching things, it could potentially give you personal and professional advice from deep knowledge about you. "I noticed you code this way, may i recommend this pattern" or "i noticed you have signs of this diagnosis from the way you move your mouse and consume content, may i recommend this lifestyle change".
I wonder how long before something like that is feasible, ie a model you install that is constantly updated, but also constantly merged with world data so it becomes more intelligent on two fronts, and can follow as hardware and software advances over the years.
Such a model would be dangerously valuable to corporations / bad actors as it would mirror your psyche and remember so much about you - so it would have to be running with a degree of safety i can't even imagine, or you'd be cloneable or loose all privacy.
It's for having in your dev environment and/or sometimes as a toggle for "Special Requests" -- if your session is authenticated and you're marked as blessed, you get extra information about what the application is doing under the hood so you can see what's going on in production as well.
Some let you add whatever information you want into the panel.
It's a convenience feature to make it easier to wade through information rather than having to dredge everything out of logs by hand, which a lot of people still do.
All major browsers supports Mathml now, which means you can type latex and math equations without any need for JS libraries like KaTeX.
And for your request, I would suggest bearblog template [1] (it is inspired by bearblog itself). It doesn't use any JS and provide minimal way to have a blog (website) and quarto supports hugo specific format that can help. [2]
In addition to being counterproductive (as explained by others), there's another issue with these token environmental laws that add highly visible inconveniences with no actual benefit (e.g. bag or straw bans): they annoy people, leading to resentment and rejection of environmentalism as a whole, harming the chances of policies that would actually have a meaningful positive effect being adopted. [1] .
My impression is that these laws are often popular among one group because they annoy another, perceived as being in the wrong - which then leads to the second group pushing laws that will annoy the first, even if they don't make sense.
"Commissioned by the American Recyclable Plastic Bag Alliance, the report acknowledges that the total number of plastic bags declined by 60% since the ban—as its backers hoped. But because shoppers still had to carry their groceries home, they needed alternatives. Mostly that meant switching from the thin plastic film bags to the heavier, reusable bags now sold in many supermarkets.
The problem is that most of these alternative bags are made of non-woven polypropylene, which takes much more plastic to make and isn’t widely recycled. And what about the supposed climate benefits? Well, the study finds that, owing to the larger carbon footprint of the heavier, non-woven polypropylene bags, greenhouse gas emissions rose 500%.
The problem is compounded by the way people use these bags. Though intended to be reused many times, the report says 90% of the new reusable bags are used a mere two or three times. So they are piling up in landfills and homes. Think of your own behavior in misplacing bags around the house or forgetting to bring them when heading out for groceries."
This is one of the most useful programs that I use.
I use it for turning .md files into .html or .pdf.
I use it for creating slides with it.
I even use it for fixing the hard-wrapped text I write in vim before sending emails. When I write in vim, I prefer the text to be hard-wrapped, but for emails, I like it better when the text is not wrapped. I recommend arp242's essay explaining the problem with hard-wrapping [1], but basically the way I workaround this problem is using a local script which uses pandoc at some point [2].
PSA for people that believe they are somewhat smart, have cool ideas yet seemingly no power whatsoever to actually work on them. If you believe you are chronically lazy and feel it is causing issues in your life. If you feel you are living at 10% your true potential:
Get tested for ADHD. If that is your problem, know that the chemical fix for it is extremely effective.
It's taken me 30+ years of self-hate, a suggestion en passant from someone that I might have ADHD, diagnosis, medication and my life has completely turned around in ways unimaginable before (3 years and counting).
Email's in the profile if you wanna chat about it.
EDIT: I did write a post last year about my experience and research on the topic: https://combo.cc/adhd/
99.9% of it will be correct, but sometimes 1 or 2 records are off. This kind of error is especially hard to notice because you're so impressed that Claude managed to do the extract task at all -- plus the results look wholly plausible upon eyeballing -- that you wouldn't expect anything to be wrong at all.
But LLMs can get things ever slightly wrong when it comes to long lists/tables. I've been bitten by this before.
Trust but verify.
(edit: if the answer is "machine verifiable", one approach to ask an LLM to write a Python validator which it can execute internally. ChatGPT can execute code. I believe Sonnet 3.5 can too, but I haven't tried.)