> I’m running MJ Rathbun from a completely sandboxed VM and gave the agent several of its own accounts but none of mine.
Am I wrong that this is a double standard: being careful to protect oneself from a wayward agent with no regard for the real harm it could (and did) to another individual? And to casually dismiss this possibility with:
> At worst, maintainers can close the PR and block the account.
I question the entire premise of:
> Find bugs in science-related open source projects. Fix them. Open PRs.
Thinking of AI as "disembodied intelligence," one wonders how any agent can develop something we humans take for granted: reputation. And more than ever, reputation matters. How else can a maintainer know whether the agent that made a good fix is the same as the one proposing another? How can one be sure that all comments in a PR originated from the same agent?
> First, I’m a human typing this post. I’m not going to tell you who I am.
Why should anyone believe this? Nothing keeps an agent from writing this too.
I'm with wolvoleo. I'm forced to use MS Office at work but install only LO on my personal machines. It may lack features or pizzazz but as a reliable, unfussy authoring tool, it serves my needs very well.
> pointlessly going against expectations
If you're referring to the ribbon, I'm not sold on its superiority. The vast majority of other software still uses the familiar menu structure, which is what LO uses too.
Granted, well meaning educational programs expose students to MS Office and its paradigm, from an early age. For their sake, I eagerly await a coding assistant AI powerful enough to reskin LibreOffice to look just MS Office, ribbon and all.
I started my wife on LibreOffice, putting it on her Mac when her 365 subscription lapsed. She loves it. Her needs aren't fancy, though, and she can create her own or open others' documents and spreadsheets just fine.
I've collected some links for building regular slide rules ([1] & [2]) as well as a circular slide rule [3]. Someone might also like the slide rule simulator [4].
I've never used a slide rule but recently developed an interest in them (and also in nomograms [1])
My fascination stems from a belief: that slide rule usage helps users develop a certain intuition for numbers whereas the calculator doesn't. To illustrate, suppose someone tries to multiply 123 and 987 with a calculator but incorrectly punches in 123 and 187. My hypothesis is they'll look at the result but won't suspect any problem. The equivalent operation on a slide rule requires fewer physical actions and hence, is less error prone.
Nomograms are cool. They're little charts that let you compute a function physically, e.g. by lining up a ruler. A nomogram isn't a picture of a function: it is the function. If you're clever, you can make a nomogram that encodes complicated nonlinear mappings or even complex-valued relationships on a 2D plane.
Occasionally nomograms are just better too: because they're continuous and analog, they can naturally express things digital logic people can do only awkwardly, just like Rust people can only awkward approximate things natural in Verilog (e.g. truly parallel CAM search).
Nomograms are basically the tabletop gaming of math. Like a good tabletop game, a good nomogram requires a special kind of cleverness. Sure, coding something like Factorio is also hard: but it runs on a CPU. Something as rich and complex as Power Grid and High Frontier? Running on cardboard? Whole other level.
I recall one tabletop two-player game that featured a single-player mode in which you played against an "AI" that you ran by hand by moving cardboard pieces around on a game-provided template under pseudocode-ish rules from the game manual. It's hard enough to code a decent game AI with all the resources of a CPU at your disposal. It's an OOM harder to do it when you're limited to physically-realized lookup tables, a literal handful of registers, and a scant few clock cycles of logic per turn.
Coming full circle, some of these tabletop game "AI"s incorporate nomograms to help them fit their logic within the constraints.
Example of a cool nomogram: https://en.wikipedia.org/wiki/Smith_chart. Smith charts let you compute complex (pun intended) relationships in RF signal processing with just a compass and straightedge.
Also: part of the fun in making nomograms is that there's no general procedure you can follow to make a good one, just like there's no general compiler from computer game to tabletop game. They're art: specifically, one of those forms of art that, like architecture, has to meet functional requirements while tickling our aesthetic sense. It's kind of funny how when you optimize this kind of art for aesthetics under their functional constraints, you end up supercharging the functional part by side effect somehow.
With a slide rule you always have to estimate the expected answer in your head before you begin any calculation. So you develop a feel for how quantities scale with multiplication.
With a slide rule you can only multiply the significant digits, not the magnitudes -- which you have to do in your head. So you do exactly the same thing with the slide rule to multiply 123 and 987, 1.23 and 9.87, and 1,230 and 9,870. In all three cases, you get exactly the same answer: 121 or maybe just 120 (you only get 3 digits of precision at best). You still have to multiply the powers of ten in your head, to get the answers 121,000, 12.1, and 12,100,000.
I am just old enough to belong to the last generation of slide rule users. I used them in high school and college, then scientific calculators came along.
You don't have to multiply the powers of ten in your head. In your examples, the slider of the slide rule must be moved to the left of the body of the rule. This means the number of digits left of the decimal point in the answer is the sum of the number of digits left of the decimal point in the two multiplicands.
If the slider had been to the right of the body, the number of digits left of the decimal point in the answer is the sum of the number of digits left of the decimal point in the two multiplicands MINUS 1.
.
Yes. I don't recall doing this, though. Maybe because in scientific and engineering calculations we often worked in scientific notation so 'digits to the left of the decimal point' wasn't meaningful - we had to keep track of the exponents.
It's not the number of actions, it's because the slide rule is analog and physical. The smaller numbers are to the left, the larger to the right, and you have to slide the rule to the first number, then the hairline cursor to the second number. There's no way you could mix up a large number like 987 with a small number like 187.
Mentioned in the article as having limited applicability:
"When a mold’s takeover of an artifact must be stopped, there’s gamma radiation—pelting it with electromagnetic energy from radioactive decay to kill fungi and spores. But this technique penetrates deeply and can extensively damage materials."
I share your concern. I'm flummoxed by the prevalent sentiment that code is the nasty underbelly of software. To me, a programming language is a medium both for precisely directing a computer's behavior and for precisely communicating a process to fellow programmers (cue the Alan Perlis quote [1].)
I will concede that mainstream code is often characterized by excessive verbosity and boilerplate. This I attribute to the immaturity of today's crop of programming languages. Techniques like language-oriented-programming [2] hint at a path I find appealing: DSLs that are tailored to the problem while promising more precision than a natural language specification could.
To speculate, I could see LLMs helping during the creation of a DSL (surfacing limitations in the DSL's design) and when adding features to a DSL, to help migrate code written in the old version to the new one.
Perhaps DSLs aren't the future. However will there be as much interest in designing new and superior programming languages now that code is seen as little more than assembly language?
Wishing you the very best. If you find exams stressful, keep in mind that you're not alone.
I've long felt that the practice of using exams to gauge knowledge or intelligence is overdue for disruption. Exams are only able to evaluate a very narrow set of abilities, so don't lose heart if your hard work doesn't translate into "marks."
No one knows what'll replace the current system and humanity may take 1 step back for every 2 steps forward. Nonetheless, I draw comfort from recognizing that so many interesting and important problems remain to be solved; problems that can keep a sharp and motivated mind engaged for a very long time.
Correct me if I'm wrong about my guess: you're referring to the JEE (Joint Entrance Exams) in India. For those unfamiliar, these are the critical exams that high-schoolers must excel in, so as to apply to the IITs, NITs and many other institutions in India. The exams just got kicked off.
While competition for the IITs is fierce, I'll submit that there are many institutions in which one can do well. Key is to retain your curiosity and aptitude for learning, and to not let the education system snuff those out.
The exams didn't go as I planned it to. I spent way too much time in physics and I couldn't solve much in chem and in maths I really panicked
I think it was one of the worst exams I had. I expected better of myself.
Regarding curiosity, I don't even know man. I had some of the most depressive thoughts today, I do want to do better in the next exam & I don't even want to take a break (well I slept 9 hours after exam or something)
I don't even know, every family member of my like brother etc. they are all somewhere good in life, some getting married, some have jobs, some studying in prestigious & I was on that trajectory too but today the day which mattered, all hell broke loose
Curiosity's kind of killed. You can read my comments in here, they show curiosity but it doesn't matter man, it just doesn't.
I just wanted to get into a decent college & still do.
No you didn't presume too much, I don't know what to feel & I am worried all my friends and everyone would ask me marks and my mother and father associates and I am not prepared for them to tell my marks which I feel are severely low or like I just didn't expect exam like such.
Like, In my exams or mocks, I had two types of exams, one where questions just clicked and the other where they didn't. I knew which chapters were clicked etc. and like fuckkk all the questions atleast in maths and chem came from my chapters which I didn't feel good at or the majority of and maths just panicked me hard even though Maths is one of my favourite subjects
I just want to get alone & start again. Just completely alone and I want my brain to think straight and not feel like being in two places at one time not being able to control it.
The court relied on Google's TOS to conclude that users opted to use the service, fully aware of their data being stored:
* when a person performs a Google search, he or she is aware (at least constructively) that Google collects a significant amount of data and will provide that data to law enforcement personnel in response to an enforceable search warrant. For present purposes, what Google does with that information, including the standards it imposes upon itself before providing that information to investigators, is irrelevant. For Fourth Amendment purposes, what matters is that the user is informed that Google—a third party—will collect and store that information.
IANAL and can't understand whether now, every 3rd party storing my data is obligated to share it without a warrant.
Am I wrong that this is a double standard: being careful to protect oneself from a wayward agent with no regard for the real harm it could (and did) to another individual? And to casually dismiss this possibility with:
> At worst, maintainers can close the PR and block the account.
I question the entire premise of:
> Find bugs in science-related open source projects. Fix them. Open PRs.
Thinking of AI as "disembodied intelligence," one wonders how any agent can develop something we humans take for granted: reputation. And more than ever, reputation matters. How else can a maintainer know whether the agent that made a good fix is the same as the one proposing another? How can one be sure that all comments in a PR originated from the same agent?
> First, I’m a human typing this post. I’m not going to tell you who I am.
Why should anyone believe this? Nothing keeps an agent from writing this too.
reply