cainxinth

There was an HN post recently about “startup anti-patterns” [0]. And the broad takeaway was that most failures stem from preventing facts on the ground from reaching you, i.e., the company plows ahead with a strategy that misses or actively ignores countervailing data.

[0] https://www.itamarnovick.com/intro-to-startup-anti-pattern-s...

I just started 1Q84. I’m only on chapter 7. It was recommended to me by a friend who is a big reader, and the last book this person recommended was Parable of the Sower, which absolutely floored me, so I’ll read anything they suggest. That said, 1Q84 is a very different book. I definitely feel an ominous tension growing, but the pacing is very measured and unhurried. I haven’t decided if I’m liking the book or not yet.

These are all accurate and bad, but I have to wonder how many of these issues are just our current “ai can’t do hands” that will eventually be solved.

I’ve read so many posts like these on hn, but I just don’t relate to them. I have loved internet access since my first dial in to a BBS in the 90s. I’ve been cruising the information superhighway at top speed every day since. There is always something new to discover.

They keep announcing new 3.8 variants. I wonder why they still haven't updated 3.1 Pro yet.

I guess that means you are officially the creator of a "classic" LLM test. Congrats!

> We are not just loosing our voice or identity we are also loosing our thinking capability.

They bolded that section. I’m not saying we need to give over all thinking to AI, but it is helpful for spotting typos.

Words are cheap, and Altman is a salesman. He’s just staying whatever he needs to say to keep people off his back. I don’t put any stock into this statement whatsoever. He may slow development, or he may not, but whether he does will depend almost entirely on what’s best for him and his company, not humanity.

It can’t do what he’s trying to do. It can’t one shot a giant project. Not reliably. It still can’t. Not even Astra. Not even close.

You need to use LLMs to build the individual components and then put it together yourself. The human architect is still needed.

Just saying: “build this complete project” is not architecting. It’s more like wishing. You will find rare examples where someone’s LLM wish came true (more or less), but I think most of these people are just burning tokens.

It's the same story every time OpenAI or Anthropic releases a new model. They are generous with compute for the first few days, and use maximum fidelity with uncompressed weights. Everything runs at its best to make a good first impression. But eventually they pare things back and the models perform a little worse.

I’m not an expert at body language, but I think anyone can see she is trying really, really hard to appear friendly and trustworthy. But when someone is trying that hard to convince you they are benign, it makes you suspicious.

I find the whole Skills.md idea a little silly. We have a technology that struggles to stay within rails due to its inherent makeup. And you think you can fix that by just telling it to?

It makes as much sense as the so-called “humanizer” tools that purport to make LLMs stop using their well-known verbal tells.

All you are doing is saying: “Hey you know that thing you can’t stop doing? Can you stop doing that?” The machine will say “Absolutely!” but eventually start doing it again.

My suggestion is to get a plain old bicycle (not an electric or stationary bike). One hour outdoors on a bike clears away at least a day’s worth of mental detritus for me.

“It’s better to be lucky than good.” - Lefty Gomez

Take a random essay and add in a bunch of the phrases that LLMs love like “load-bearing,” “crucial,” structural,” and “woven,” and then submit the original and the edited version to an LLM and ask which is better. It will choose the second one virtually every time. They have ingrained biases that associate those words with good writing and arguments.

I looked closely. They are going the wrong way. Still the best one I've seen by a lot.

> Pompous things like "Your keys, supercharged" or weird yoda-speak stuff like "searches the app remembers"...

It's copywriting. They fed these models the internet, which is loaded with it.

Somewhat related, I’ve long wanted an ereader app that color coded the characters in a novel.

> It's exactly like vibe coding but for computer tasks.

That’s precisely why I don’t use it. LLMs are still not reliable enough for me to trust them with my actual system.

This is how I proofread. I play a screen reader while I read my work. The brain does some autocorrecting that you don’t even notice and occasionally skips over a typo when you read it without the audio.

Has anyone found a useable fake review detector since Fakespot and Review Meta went down?

I also tracked my clothing for three years, but not to evaluate expenditures. I recorded what I wore while cycling along with the weather conditions that day and notes about how my clothes performed. When I want to go for a ride, I check the weather and then check my spreadsheet for a day with similar conditions.

I'm fascinated by people who get Type 1 fun from reading the classics. I get type 2 at best.

I virtually never click anything texted to me (other than personal stuff from friends and family). If a bank or a shipping company texts me, I go to their website and look up the information myself.

Like the author of this post, I also have a better than average eye for spotting scams, but it’s foolish to assume you’ll be right 100% of the time.

>As for semicolons, semicolons are neat, they are like the coolest punctuation mark after the em dash — the fundamental indecision compressed into a single punctuation mark — you are neither starting a new sentence nor you are, exactly, continuing an existing one. You are just… piling your stuff together.

A semicolon is more than a comma but less than a colon. It's for when I want to use a comma, but it would be grammatically incorrect to.

Learning that MIT has an ancient bottle of rotting milk reminded me of my 9th grade science teacher. He had been teaching for decades and had a tremendously messy little office attached to his classroom. The pièce de résistance of his menagerie of scientific oddities was a large fish tank filled with many strata of mottled goos. I asked what it was and he said it used to be an ordinary aquarium, filled with fish and water, and he just left it alone for 20 years.

> It’s a classic small business supplier story: the big guys will do what works best for them and the little guy will have to deal with it in order to get the benefit.

“The strong do what they can and the weak suffer what they must.” - The Melian Dialogue

When my niece was very young, I showed her my Polaris pool cleaner. It’s a simple contraption. Just some hoses attached to a small box with a few wheels. It rolls around the pool, sucking up leaves on the bottom.

She took it all in, and then turned to me and asked: “Is that robot’s family living in the ocean?”