Hacker Newsnew | past | comments | ask | show | jobs | submit | bananaflag's commentslogin

Very good!

This is something that existed in the past and I used successfully, but services like this tend to disappear


That’s a completely fair concern. Services in this category do need to earn trust over time. I built the backend to handle a fair amount of traffic, so I’m not too worried about growth on that side. My goal is definitely to keep this running for the long term, not treat it like a one-off project

> Always has been, and if you paid attention in CS class, you know the limits of those things.

I don't remember ever learning a theorem stating that computers cannot surpass humans.


Haha that's not what the post (or the post it links to) says. Every CS student should know there's no free lunch in search and optimization. There's tradeoffs between random search, evolutionary algorithms, and convex optimization. https://en.wikipedia.org/wiki/No_free_lunch_in_search_and_op...

There's an AI "smell" to things that are generated. Why is that? Mode collapse is impossible to see from a small number of samples. Are we mode collapsing society? How would we know if we were?

Also, will computers surpass humans has such an implicit bias in it. Have humans surpassed ants? Have ants surpassed rocks? Have jet planes surpassed teletubbies?


> Every CS student should know there's no free lunch in search and optimization

The no free lunch theorem is so absurdly limited because of the constraints that it's IMO a tautology and fundamentally irrelevant outside of exceptionally tiny areas. You can't have one search algorithm that's better than others on average when searching entirely random things with no structure? 1. Yes, obviously. Nice to have a formulation but it's not exactly a surprise and 2. That's not what we deal with in the real world.


I know for a fact StanleyNickels™ have surpassed, nay, exceeded SchruteBucks™

I get what you're saying, but I remember watching teletubbies back in the days with my nephew, and all questions of the form:

Have ____ surpassed teletubbies?

Can always be answered in the affirmative.


I remember P vs NP

Unfortunately, substack paywalls cannot be bypassed



Scott took it too literally. See also how the broader rationalist community took issue with Sam Kriss for inventing a not-obviously-fake historical figure.

The biggest takeaway for me is that you shouldn't expect to succeed as a manager by meeting (or exceeding) KPIs. It's about as effective as being a "nice guy" and expecting intimacy in return.

The KPIs are there for assigning blame, not for identifying key personnel. You can game them to increase your compensation if you are already doing something that an even bigger manager finds useful and important. Conversely, you can get away with half-assing every official performance indicator as long as you keep delivering the real thing.


That’s a good takeaway and if anyone doubts you just think about how you set “goals” in the HR system every year during annual review time , vs. what your boss talks to you about

> Scott took it too literally

He does say in section (I):

> I was particularly told to “take it as literally as possible”


He shouldn't have taken this advice literally either.

Just curious what’s the definition of “success” here? Getting promoted and getting a better compensation?

And self-actualization, if you're into that.

> arrested development is the dark side of strengths in the sense of Positive Psychology

I see some correlation here to hesitancy in adopting LLMs for coding.


Explain.

to hazard a guess: the better your coding model is, the less you have to assess your fundamentals, and thereby suffer arrested development.

I was thinking of our natural reluctance to adopt newer and better tools because of our comfort and expertise with old ones.

I know I should have experimented with LLMs sooner, but leaned into my instinctive "VIM has gotten me this far" attitude.


Liked this comment:

"If we could convince [any] Sociopath that we were all Losers, we might be able to entice them into spilling their secrets as 'Straighttalk'. (Arguably that's what this book is..)"

On one hand Rao doesn't say much about Gametalk (he basically defers to Eric Berne) which is the Loser's sociolect and should well be our default.

On the other, Rao much more optimistic than Orwell, who declared doublespeak the lingua franca?


> On the other, Rao much more optimistic than Orwell, who declared doublespeak the lingua franca?

If time travel were possible, one of the first things I'd do is introduce Orwell to the 'algospeak' of today. This would do two things, firstly it'd show him a decent piece of evidence that Newspeak isn't as effective a tool for limiting human thought as he believed, and secondly he'd have to write another version of Politics and the English Language aimed at the language sins of attention economy era social media.


and then I'd show him a news broadcast from last week, where the president of the United States of America literally said "War is peace".

Is the use of "literally" here, and the use of quotes, meant to be taken literally (as in, he literally said this)?

Or is this the sense of "literally" which actually means "figuratively"?


I love that we ruined our escape character in language.

We now need a new way to convey to listeners that our words need to be taken literally. Perhaps the VBA method and just double the 'literally's:

I took a flaming shot of tequila and my mouth is literally literally on fire


A post to the Truth Social account for Donald Trump included: "The heavy and pinpoint bombing, however, will continue, uninterrupted throughout the week or, as long as necessary to achieve our objective of PEACE THROUGHOUT THE MIDDLE EAST AND, INDEED, THE WORLD!"

That's the closest thing I'm finding. Not seeing reporting that he literally said "war is peace".


That sounds more like "peace through superior firepower" rather than "war is peace".

Can we be literal? It means peace through using superior firepower to kill people.

Sure, but it's not equating the states of war and peace, but asserting that war is a method for achieving peace, presumably when everyone on the other side is incapable or undesirous of attacking or threatening same.

Are we forgetting the context of who started the war?

It's just not relevant to the statement "I'm punching until they can't punch back" who threw the first punch.

I'm not sure I follow.

What I mean is that it makes no sense to say you're fighting a war to achieve peace if you're the one who started the war, breaking the peace that existed before then.

If peace is your goal, then the status quo was already in line with your goal. Starting the war contradicts your stated goal of peace.

Do I really have to spell all that out?


There's a vast gulf between "having" superior firepower as a deterrent and "using" superior firepower for mass murder, particularly against elementary schools and desalination plants. The latter is war, at its worst.

Beatings will continue until morale improves?

You can't make this up.


The Berne books Rao cites as explanations of Gametalk are solidly good entries in of themselves, although it's probably best to use an LLM to get search results of the best introductions to TA first to see if they've been surpassed.

Adhering to the predictable/ritualistic/comfortable nature of "Gametalk",

Here's one question I asked:

"How does Eric Berne's Gametalk as interpreted by Venkatesh Rao signal to the sociopaths that those who engage in them are losers worth talking to? Distinguish between "channels" that Eric has identified as well as new signals that Rao or others have discovered."

https://youtu.be/9B3oem_56jg?t=52s


Can you expand on your included youtube link? It's not clear the relationship.

I'll admit the connection is loose, personally found it amusing because:

Mike is the archetypal nihilist (Sociopath or Loser), the other two would potentially be engaging in a Clueless interaction if Mike wasn't there, according to the Scott/Rao theory of jokes, you need 3 for a Loser joke.

The preceding banter seems to be more of a Loser Gametalk: no social status is at stake; it's irrelevant to their white-collar role. Mike's Straighttalk intervention is typical of a sociopath; the wall breaking joke is that these Losers don't know what his real job is. If they did, the pointless but playful debate would have died a violent death-- because it'd get too real

If these were Clueless middle managers debating their value to their company, it might even be out of character for Mike to notice them..


I guess one day there will be a massive leak of executives chats with their LLMs, and we'll find out what they really think.

I think that’s called the Epstein Files.

Used to think that Epstein was a Posturetalker but turns out he is a native Gametalker

Yeah, and I always find the phrase "publish in a conference" to sound vaguely oxymoron-ish.

Well, it's published in the conference's proceedings, and presented at the conference.

https://en.wikipedia.org/wiki/Proceedings


> This is exactly the same plot as the first three movies

Well I hoped with each one there would be a different plot.

The only thing essentially different is the Reylo plot which is kinda spread between the three movies. It makes for like half of an original movie.


I've been using RSS daily since 2008 (on feedly since 2013)

Does your Mother use RSS daily?

Does your kid?


No, but I use RSS instead of checking every single website compulsively, which people generally don't do (I barely know anyone irl who follows a blog, RSS or not).

So it's not a problem with the RSS tech, but with its use case.


I think everybody knows about Left Behind (but nevertheless I'll mention it for good measure), but if you want a shorter fictional version of "this is what these people actually believe", check out the Christ Clone Trilogy by James BeauSeigneur. (Left Behind is too long and poorly written, I myself got bored at like the 8th book or so.)

Can't you just buy it from the grocery store?


Yup, saw that review. My takeaway is if one has knowledge and training level of Scott Alexander then this book has nothing new to offer. But since most folks don't so this maybe a interesting read.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: