And even the headline gets it wrong. They asked 1,000 people, over half of whom say they lacked basic statistical understanding.
The first question was:
“How much do you understand about statistics and p-values?”
not probability, p-values.
I had to look up p-values to check it meant what I thought it did to avoid making a fool of myself in this comment. I knew I was probably right, but possibly wrong.
The same goes for my assumption that this survey was engineered to deliver a bad result.
Why? because it makes it easier to get funding for a the author's solution to the problem. Again, probably right, possibly wrong.
Statistics, like math, is a language, first learn the vocabulary, then you can judge the meaning
So what is the mechanism that the administratuon would use to block access?
I get how China could block access to Chinese LLMs, but any startup that can't or won't bypass a US imposed block on a Chinese offering was going to fail anyway.
This makes complete sense in an environment where people transition from noob to senior engineer within the same company.
It makes less sense in an era when tenure is better measured in months than years.
It makes even less sense in an era of LLMs.
One area where it might be relevant is the military. People are more likely to stay for longer (unvalidated assumption) and the same personnel jacket follows them if they are transferred.
It might also be thought of as a guide as to when to jump ship. If you have managed to get yourself categorized as a C, then leave. Start fresh somewhere else, take the learning with you, and discover if you have what it takes to make it as an A or B.
I would argue that out makes even more sense in the era of LLMs. LLM shaped tasks are tasks that we would hand out to junior engineers. Now, I can implement one of these tasks in 1hr instead of waiting for a junior engineer to finish this in 1-3 days. This means the equation for investing in junior engineers has shifted towards disfavoring investment.
This is my own personal experience as a senior engineer.
I hired my first ever intern for myself this past week for a personal. Rather than expecting someone who is experienced to knock out tasks or whatever & try to justify the expense, Intern and I just check in once a day about whatever feels like most important, and each do our thing. We chat as needed. I told Intern we’ll work on whatever, just making sure that they will have something tangible and targeted to show at the end of the summer. No ticket tracking or Slack or anything. Just texting, video chats, and the occasional email. I pay someone to listen to me rant long enough to get to the point, giving me focus that’s hard to find on my own, intern works on targeted things for a day at a time, and we’re just plodding along. It’s great and I find the process to be extremely refreshing.
When I’m trying to brainstorm with Claude Code or pick-an-AI-tool, I find the process frustrating and draining. The results I get from trying to do everything myself with a robo-junior are mediocre and uninspiring.
I didn’t even interview intern. I just reached out directly on LinkedIn and offered a summer internship. I figured anyone with a half decent profile will be smart enough to follow along and offer ideas of their own. Basically my thought was if I offer ever and expect nothing, I’ll take all the pressure off, and just let them work. Ask me again in a months if this was a good idea or not.
I think working with newbs is fantastic, and I plan to do a lot more of it.
> It makes less sense in an era when tenure is better measured in months than years.
If a person's tenure in companies is measured in months then they're signalling they're a C by your logic, or is at least raising a red flag to whoever's hiring. I may be showing my age but that sounds wild to me if that's the norm now.
Maybe this depends on the framing? Ex. 18 months is fairly common in some circles, but that could alternatively have been expressed as a year and a half.
6-12 months? Red flag
12-24 months, especially early-to-mid career? Not uncommon
> It makes less sense in an era when tenure is better measured in months than years.
Is this common though? Most places I have worked have had people with pretty long tenures. Maybe Silicon Valley during peak ZIRP where you could just keep jumping as long as you could pass the leet code... but then why aren't you staying for RSU vesting?
It would have been interesting to see if there was any difference relating to CYP1A2 (Cytochrome P450 1A2), the fast metabolizers and the slow metabolizers.
> The guys out there with big Che Guevara energy are the real ones building and perpetuating a misery machine fueled by your ideology and nothing else
---
Ha, if the author hadn't mentioned that he was in South Beach on vacation, these lines would still make me think, "Here's a guy who sounds like he's in South Beach but is definitely not from Miami!
This also works for kitchens.
What is most interesting is that it begins to impact what one buys. It turns out that, after a decade or so, one can predict which 'must have' gadget or appliance is actually just a very seductive dotless wonder.
Honestly I think kitchens have way less stuff than a lab like this and following Alton Brown's advice of avoiding "unitaskers" will basically guarantee you don't get overwhelmed.
The unitasker advice is also a bit difficult for inexperienced people to follow, from what I've seen. A stand mixer is a great multitasker on paper, whereas a speed peeler does exactly one thing. Yet the latter will be used massively more often than the former in most kitchens. Probably the most used tool in my kitchen (after knives and cutting boards) is the kettle, another unitasker.
I generally agree with that and there are some small kitchen appliances I don’t use often but when I need them I need them. Funnily an electric drip coffee maker is one of the things I somewhat regret replacing after a kitchen fire but I think insurance and may be handy for company at some point.
Some problems with that: "discredited" isn't a binary; memes (in the original sense) from papers circulate easily even if nobody pays the publishers (the problem is arguably even worse, because people will read abstracts for free and distort the claims further); there might not be higher quality alternatives.
This also applies, on a larger scale, when one adds data to a medium like a sheet of paper, the graphite or ink adds to the mass of the storage medium.
But does this constitute data?
The maximum mass would be achieved by covering the entire sheet with graphite/ink which, it could be argued is not data (unless you consider it to be a binary cell in a larger byte of data).
I don't know the physics of thermal paper, but I suspect that it might be the opposite.
My point?
This is not evidence that data has mass, it is evidence that transcribing data onto a storage medium may change the mass of the storage medium, and that change maybe positive or negative.
Perhaps I should have this carved on my tomb stone...
The first question was: “How much do you understand about statistics and p-values?”
not probability, p-values.
I had to look up p-values to check it meant what I thought it did to avoid making a fool of myself in this comment. I knew I was probably right, but possibly wrong.
The same goes for my assumption that this survey was engineered to deliver a bad result. Why? because it makes it easier to get funding for a the author's solution to the problem. Again, probably right, possibly wrong.
Statistics, like math, is a language, first learn the vocabulary, then you can judge the meaning