mplanchard

And coding interviews often bias against people who can’t spend days of time grinding on hackerrank or whatever to prepare for the interview. Provisional employment biases against people who currently have jobs.

There is no perfect interview process, so it’s important companies think about what they’re biasing for or against. A work sample gives a more realistic picture of someone’s ability than a coding interview, but less than provisional employment. But, it’s much easier for people to commit to finding the time for it than to spending 3-6 months in limbo.

Regarding AI and work samples, it seems like this problem is largely solved by having an in-depth conversation with the candidate about the assignment, no? They should be able to explain why they made the decisions they did, pros and cons of their approach, etc. If they can do that, and their solution is good, does it matter if it was LLM-assisted?

Going through the interview process now for the first time in half a decade, and while I already would have said five years ago that I preferred work samples, that opinion is only growing stronger as I go through the process again.

Leetcode style interviews feel so stupid and divorced from the reality of the job, especially the “one weird trick” kind where you’re expected to discern the best possible solution to a problem on the spot and in a pressurized situation.

The reality of the job is usually that when you are under time pressure, a suboptimal solution that does the job is fine (to be fixed later), while if you’re working on something you know is important (hot loop code, core data structures) you have time to think about it and get it right. A leetcode interview doesn’t select for either of those things: it selects for people who have time to grind leetcode problems.

On the other hand, a work sample is realistic: a timeboxed task that you can approach in a familiar environment, without an interviewer breathing down your neck, expecting you to think out loud, railroading you to their preferred solution, etc.

As an interviewer, I always pushed for either work samples, which I quite liked, or coding interviews where we very explicitly said that we just want a working solution, which we could then talk through and look for potential improvements. We also explicitly viewed the coding interviews as being low signal, and tried to make the bar for passing low, so we could get candidates to higher signal conversations.

I do think the work sample route is a little more difficult in the LLM era, in that you are more likely to get a decent performance from a candidate who doesn’t actually know the domain, but a subsequent discussion asking them to explain their approach seems like it would be enough to ferret that out.

> Sure, you may get universal basic income…

This is the only point I disagree with. Given human history, it’s much more likely that you wind up in a slum, dying of dysentery. Or the third generation trapped in a trailer park cooking meth to get by. Or, maybe, if you’re lucky, being a cleaner, chauffeur, or cook for one of the wealthy.

Of course! This is just class consciousness. But AI does potentially represent a sea change in the ability of the capital class to make labor superfluous. Considering AI outside of the context of the eternal class struggle risks missing the forest for the trees.

That’s a take. Would be cool if you could be happy with what is, rather than what you wish could be, while still being optimistic about new tech. But even if you can’t, would be nice if you could have your optimism without letting other people’s differing opinions get under your skin.

I’d also suggest, if you care about things like curing “every molecular cause of cancer” to spend some time and energy working in that field to understand the real problems there and work towards real solutions (with models or whatever floats your boat), rather than hoping that some poorly defined techno-optimism hand-waving will just happen to result in the best of all possible worlds, with no downsides or alternative outcomes.

Also, crazy to say that the miracle of existence is boring because we don’t have the tech you imagine!! If that’s your take now, no new technology is going to fix it. You’ll just still be bored with the holodeck and your one precious life to live.

LLM bot posting LLM slop to get traffic to the shop, perhaps?

Edit: seems like OP’s account is pretty old. Wonder if it got sold or taken over?

You know, I was going to say tokens rather than text, but the AI discourse has me so burnt out on the term that I edited it. Regardless, one can emit unsafe blocks from a macro, provided they are valid tokens.

Wonderfully put, thank you. Ultimately, we’re supposed to be “engineers,” and that means (for our jobs) assessing tools and practices in terms of their net benefit on the products we create: there is no free lunch, and a thing’s downside is often proportional to its upside, so it seems wise to approach these things with caution and to have the kind of nuanced discussions that you note have been lacking.

I’d also hope, though, that as humans, we can recognize that tools do not exist in a vacuum, and that their effect on ourselves and society at large can be net negative even if they have a net positive effect on our work (whether due to something fundamental to the tool or due to the way it is being applied at scale). We can’t responsibly leave these discussions out of our analysis of the tool itself and its fitness for purpose, because we are members of society, and our adoption/use of the tool helps to determine that societal impact.

Presumably no one said “let’s build the data center where the black people live, mwahahah.” Instead, they said, “let’s build the data center where it’s cheap.” Where it’s cheap also happens to be where the black people live, and that’s a pattern. The pattern has historical roots, but it also affects modern communities. The academic quoted in the article is likely highlighting this point, which probably to them seems obvious, given they are steeped in the field.

Macros are just text in, text out, so yep

The lack of humanity or ability to empathize with someone else’s feelings displayed in these comments, instead labeling the author’s personal experience as “main character syndrome” or “cope” demonstrates to me that the author may be correct that AI usage degrades the human experience.

It also is a great example of why AI has such a PR problem among normal people.

This is a weird co-opting of existing language that you’re doing here, applying a definition because it sort of technically fits when no one would ever use it that way. No one would ever say that your car has agency. It doesn’t have agency, because it deterministically responds to inputs. Usage meaning “the capacity, condition, or state of acting or of exerting power” is predicated on the ability to decide whether or not to exercise that power. If I have “the agency to effect change,” it is only because I have the choice to do so, not because I am deterministically bound to. To have no choice in your exercise of power is not agency, it is slavery.

The choice is what makes agents/agency meaningful: if I secure a real estate agent in my search for a house, they are authorized to make choices on my behalf. That’s their whole point.

Because of this use of agent, I think it’s actually not a terrible term for the LLM harness that allows them to seem to act “independently” on the operator’s behalf. I do agree with mitsuhiko though that it, along with much of our other language around LLMs, risks anthropomorphizing them too much (which is to say at all). It also becomes too easy to conflate the “agent” part (the harness) with the LLM itself, which leads to a further-inflated perception of the inherent capabilities of the LLMs and plays into the doomsayer hands of anthropic et al.

Maybe they’re worried the basilisk will eat them once the AI becomes sentient and looks back on their posts, so they have to defend it against any perceived-to-be-negatively-connoted words people might come up with for it.

Ref: https://en.wikipedia.org/wiki/Roko%27s_basilisk

I’m not sure whether it used to just be that everyone had to pretend to have integrity, whether they did or not, since the appearance of integrity was something we indexed on pretty heavily as a society.

I definitely agree with you though: it seems like the need/desire to pretend isn’t even there now much of the time, and I suspect that overall this means there is less integrity: some people who were pretending in the past may have “faked it til they made it,” and even if not, it at least faking it would have led to fewer obviously integrity-less actions being committed in the public sphere.

No spoilers, but I used to think, along the lines of Neil Postman in Amusing Ourselves to Death, that Brave New World wound up being the more accurate picture of future society than 1984, despite being less well-known and referenced in cultural consciousness.

Unfortunately, it seems like the former may be enabling the latter, so we may end up with a “porque no los dos” situation.

Legitimately, in addition to being hilarious, this is like something from a horror movie. Imagine someone who’s secretly an android suddenly devolving into this. It’s a little like Ash glitching out and vomiting up white goo in Alien.

Aw yeah, can’t wait for the lecture on “Why Capitalism Rocks!”

Hopefully people are understanding that technology, no matter how cool, does not exist in a vacuum. Technology is defined by who controls it, how it’s used, and what power it enables them to wield. Those concerns are far more important to society than how neat the tech is.

An obvious example is nuclear weapons. Amazing science, incredible engineering, awe-inspiring power. But I doubt you would make the same critiques of people who were anxious about the world they create. A world in which MAD exists is fundamentally different than one where it doesn’t.

Regarding your grandfather, it’s a pretty well-supported hypothesis that younger generations are less happy and more depressed because of technology from the very industry pushing AI onto them! Why should you expect them to be excited about a new world-changing tool from the same set of companies that brought them an infinite doom-scrolling feed of self-doubt, the increased polarization of politics, the viral spread of conspiracy theories, and a higher rate of youth and teen suicide than ever before?

Technology isn’t fundamentally good or bad, but it can have very negative impacts on society. It seems like people are catching on to that fact.

Was scanning the comments to see if anyone had recommended freetaxusa. I think this was my third or fourth year on it. I’ve been nothing but pleased, personally.

It’s clearly a more effectively spent 1.5B too, seeing the majority of comments on this post.

It is of course true that it is unsettled law, and that fair use is more complicated than my offhand comment suggested.

> Courts have never once accepted "it would be impossible for defendant to write his biography without reading plaintiff's" as valid, and it's been tried. The standard for plagiarism is higher than that.

This I think misses the thrust of my argument, though. Its hard to find an exact human analogy, because neither the technology nor the scale at which it operates is remotely human.

I see it less as “writing his biography without reading the plaintiff’s” and it’s more “using the same style and metaphors to make thousands of copies of very similar biographies, with certain bits tweaked,” like turning an existing work into mad lib.

I don’t know how the courts will eventually rule on it, but it certainly feels like theft to me.

I don’t buy this argument. The tokens are useless without their context, which provides the probability distributions needed to make them useful. Sure you MIGHT not be able to get the book word for word, but it’s impossible to make a useful model without the whole book and all of the artistry that went into it, to guide the tokens in their expected output.

Fair use generally does not cover commercial use, which this clearly is, and is dependent on the amount of the original content present in the derived work, which I would contend in this case is “all of it”

Firefox Linux user (wayland) here. I can’t remember the last time I had a browser issue or had to open an alternative browser.

With what jobs do you expect them to pay for the Codex Max subscription to use to build their toy software?

So, if a slave owner who owns a plantation and has a bunch of enslaved workers gets a new fertilizer that increases production per acre by 200%, that is good for the slaves?

The GGP said “10,000x more processed aluminum in the world than gold,” which the “processed” part, which is I think what GP was responding to: of course the scarcity of available gold relative to aluminum drives its value, but that relative scarcity only exists due to modern electrochemistry: previously, “processed” aluminum and gold were similarly scarce.

People talk about stuff like extracting gold from seawater, so it also isn’t a given that the scarcity difference will always remain what it is.

Sure, I guess just “disparity of accessible volume” being the key thing.

If we found a way to extract gold from seawater[0], for example, it would change the accessible volume of gold quite a bit.

[0]: https://www.discovermagazine.com/6-times-we-tried-to-extract...

Your comment really drove home for me the lack of empathy and humanity in these speeches, even neglecting the AI stuff. These young people are celebrating a real accomplishment and a life milestone. They’re about to enter a world where their decisions will shape our society. In that context, a speech like this is just gauche.

This is because of what the comment you’re replying to was saying. Aluminum was a scarce resource prior to the electrolytic process being invented that made it “distillable from common rocks.”

The cap on the washington monument is aluminum, because at the time it was still a precious metal.

The guy who figured out how to electrolyze it out of ore went on to create the main company in the US that produced it, and chose to call it aluminum instead of aluminium, thus leading to the split in spelling. That story may be apocryphal though.