khuey · HackerTrans

5 godzin temudiscuss

> And how many years of being passed over for promotions, I wonder.

He was Vice President of Product Design when he left Apple. How many more promotions could there be?

khuey

wczorajdiscuss

I don't disagree with what you're saying but I want to point out that it's rather unusual for (American) ambulances to carry blood, and probably more of them should.

https://www.redcross.org/about-us/news-and-events/news/10-wa...

khuey

wczorajdiscuss

> Back then I found the Rust community had interest and respect for Zig, so the discourse was very much one sided.

In hindsight (and at risk of starting a flame war), it's easier to be magnanimous when you are winning/have won.

khuey

przedwczorajdiscuss

If they did the opposite and called it Mygres or whatever people would give them shit for that too.

khuey

przedwczorajdiscuss

> what's a single concept in C (e.g. "the pointer") easily has a dozen specialized equivalents in Rust

That "the pointer" is a single concept in C is the root cause of absurd numbers of bugs over the history of C.

The questions of how long the memory a pointer points to lives, what threads it is safe to access from, and how it is allocated and destroyed all still exist in C. Those answers are just implicit rather than explicit like they are in Rust. It's not like Python/etc where you don't have to worry about these things.

khuey

przedwczorajdiscuss

I make no claim as to whether the change makes sense given that I didn't look at the callers of this function, but Result<bool> is an entirely reasonable pattern in Rust. If you want the callers to be able to distinguish between "has the subclass", "doesn't have the subclass", and "something went wrong" this is idiomatic Rust.

khuey

5 dni temudiscuss

That was my imprecise summary of what the contracts say. The actual language is:

"OpenAI will only use Customer Content as necessary to provide Customer with the Services, comply with applicable law, enforce the OpenAI Policies, and prevent abuse. OpenAI will not use Customer Content to develop or improve the Services, unless Customer explicitly agrees to such use."

"Anthropic agrees that Customer (a) retains all rights to its Inputs, and (b) owns its Outputs. Anthropic disclaims any rights it receives to the Customer Content under these Terms. Subject to Customer’s compliance with these Terms, Anthropic hereby assigns to Customer its right, title and interest (if any) in and to Outputs. Anthropic may not train models on Customer Content from Services."

khuey

9 dni temudiscuss

The Enterprise versions of Anthropic/OpenAI's products say they don't train on input data so either this is a misinformed rant or it's an allegation that the AI labs are perpetrating an enormous fraud on their enterprise customers.

khuey

11 dni temudiscuss

This is suburban Richmond, not a "small rural county".

khuey

11 dni temudiscuss

Claude didn't "write" anything until a meatbag told it to.

khuey

11 dni temudiscuss

The most obvious ones to me are the sentences that aren't actually sentences but just lists of stuff (and sometimes not even proper sentences). It starts at the third sentence and there's one in the majority of paragraphs. LLMs love writing like this but humans don't do this.

"Support tickets, lead follow-ups, reports, internal updates, research, data entry, task assignment, status checks."

"Real work has missing data, unclear requests, old records, broken integrations, private context, bad formatting, vague instructions, and exceptions nobody wrote down."

"They still need ownership. They need monitoring, logs, fallbacks, permissions, updates, and someone who understands both the business and the software."

"People are busy tuning prompts, adjusting workflows, adding tools, joining calls about the automation, reviewing outputs, and fixing strange mistakes."

"It can speed up writing, coding, research, support, operations, and internal tools."

"It can fail, drift, break, make bad assumptions, and produce bad output with confidence. It can depend on tools that change, APIs that fail, and data that gets messy."

"It has to be designed, shipped, watched, fixed, and improved when the business changes."

"The part where the system meets real users, real data, real edge cases, real failures, and real Mondays."

khuey

13 dni temudiscuss

Yes, there are cultural reasons crime is lower in East Asia too, but I haven't been to a major city there that doesn't have an extensive surveillance system.

khuey

13 dni temudiscuss

My experience in major East Asian cities (predominantly Tokyo and Taipei) is that they have extensive networks of surveillance cameras operated by or accessible to the police.

khuey

13 dni temudiscuss

Ford ended up paying $365 million (roughly $2200 per van) to settle a lawsuit from the government over that.

khuey

13 dni temudiscuss

The 99 in his TM99 is also an arbitrary number.

The more interesting part of the talk is how using what he calls "fenceposts" naturally results in bunching up against the fenceposts because even small regressions that cross the fenceposts are blocked but regressions of any size that don't cross a fencepost go unnoticed.

khuey

14 dni temudiscuss

https://socketsite.com/archives/2022/02/there-are-not-40000-...

khuey

16 dni temudiscuss

> You would literally be spying on employees 24x7 for weeks at a time with the express goal of replacing them someday.

Isn't this almost exactly what Zuck is trying to do at Meta?

https://www.reuters.com/sustainability/boards-policy-regulat...

khuey

18 dni temudiscuss

Having to write the "I did not cheat" pledge in cursive was the most difficult part of the SAT for me.

khuey

19 dni temudiscuss

This is an easily solved problem if you start discounting referrals from employees who refer a lot of people who generally fail your screening interviews.

khuey

19 dni temudiscuss

Applicants did this routinely ten years ago.

comments