ekidd

34 ms·

421.
by ekidd 4y ago
> Maybe Manhattan COL is insane, Manhattan COL is completely ludicrous. I knew a lawyer who lived in an actual closet ; it was exactly long enough to hold a small bed and about twice as wide. He shared a bathroom and a kitchen with th
422.
by ekidd 4y ago
> I wasn't willing to play chicken and see if they fired me. That makes negotiation harder. There a number of books from the Harvard Negotiation Project which offer roughly the following advice: - Make sure your plan B is good eno
423.
by ekidd 4y ago
> how language books are more recommended than works that intend to be foundational (like TAOCP or SWEBOK, disclaimer, I've read neither of them). I have read some of the "foundational" works, at least well enough to kno
424.
by ekidd 4y ago
> but wanted to adjust the terms in the non-compete section of their contract (something you hear suggested a lot by people here on HN). You probably hear this a lot on HN because: - In the US, programmer contracts are often negotiable
425.
by ekidd 4y ago
> it really makes me wonder if AI can really progress past a certain point when humans keep dragging it down with their own biases. I really sincerely hope that if we ever do create AI, that we teach it human biases. Like, "Murd
426.
by ekidd 4y ago
> I am sort of fearful of AI being too biased from its creators, because unlike with a search engine you can't try to find alternative outputs/viewpoints. Look at it a another way. People have already bypassed the filters and
427.
by ekidd 4y ago
Thank you, that is very helpful!
428.
by ekidd 4y ago
> Then you have that sheaves are the natural language for data fusion and sensor integration - which doesn’t apply to people who don’t know the topic, but is an industrial reason to learn it. Ah, this is super interesting. Are there
429.
by ekidd 4y ago
> Look, you don't want to use a table saw like that. For... all the common sense reasons. For example, you might have an increased risk of wood kickback, very nearly lose a leg, and spend years in physical therapy trying to replac
430.
by ekidd 4y ago
> I've long subscribed to the philosophy that you should buy the cheapest tool you can find and use safely, use it until it wears out, breaks, or your skill surpasses the capability of the tool - and only then should you spend mone
431.
by ekidd 4y ago
> blatant attempts at ban evasion This is a very strange way to describe what's going on at Twitter. Musk has loudly announced that you're not allowed to even link to your other profiles on any other social media sites, or
432.
by ekidd 4y ago
Realistically, each Mastodon instance sets its own policies. Most publish lists of the instances they've banned. When I looked at those lists a month ago, it seemed like the median instance banned 6-10 other instances. Typically, the b
433.
by ekidd 4y ago
As far as I know, the power company made the changes to their fees with the full approval of state regulators. They still provide net metering for the electricity that I buy, as per the contract. They just broke out a ~$35 monthly fee for g
434.
by ekidd 4y ago
> "I want your cashiers to buy them from me at the retail price!" It's not anywhere near that simple. Electric power prices change dramatically throughout the day. For example, even retail prices may more than double dur
435.
by ekidd 4y ago
> The chapters may seem okay individually, but you'll find chapter 5 contains statements which completely contradict those in chapter 1. I can't find documentation for ChatGPT's "context window", or how much te
436.
by ekidd 4y ago
> This is the same with all kinds of generative models. Humans are a critical part of the system. Part of what I find frustrating about Stable Diffusion (and to a lesser extent Dall·E) is that they're so ridiculously far behind Ch
437.
by ekidd 4y ago
I've seen Google Maps try to direct people down "roads" that would cause problems on a horse. And some of these errors have persisted for years, despite many reports to Google.
438.
by ekidd 4y ago
My mental model of how ChatGPT might superficially "understand" its own limitations would be because, for example, it might have been specifically trained by people who do understand the limitations of the system.
439.
by ekidd 4y ago
The US state of Vermont actually has a constitutional right to hunt on open, private lands. This dates from the original state constitution. If you don't want people hunting in your woods, you have several options. The simplest is to
440.
by ekidd 4y ago
Of course, it's entirely possible that ChatGPT doesn't understand its own limitations and inner workings. And when we ask it how it works, it just makes up a plausible answer. There is more than one cognitive scientist that argues
441.
by ekidd 4y ago
ChatGPT is performing significantly better in my personal tests than the original public GPT did years ago. The original GPT had occasional strokes of cleverness. ChatGPT has far better understanding of the questions, and it maintains sig
442.
by ekidd 4y ago
Yes. I gave ChatGPT two favorite coding screens I'd used in the past. They're nothing fancy—ordinary, everyday coding tasks. ChatGPT implemented the first program successfully, and explained how it worked. (It compulsively explain
443.
by ekidd 4y ago
Rust would not be my first choice for a simple CRUD app. There are too many other tools that handle that space well. I have used Rust at a startup, and it worked out extremely well. But we used it for CLI tools (where it shines), and perf
444.
by ekidd 4y ago
I've spent a fair bit of time running a fuzzer on Rust code that parses untrustworthy data. And the thing that saved my code time and time again was that Rust has runtime bounds checks. Even if I messed up index calculations, I'd
445.
by ekidd 4y ago
Mostly because TypeScript and typed Python are often more accessible to more developers. One of the tradeoffs of programming in Rust is that you need to distinguish between value vs. reference, stack vs. heap, and a number of other choices
446.
by ekidd 4y ago
I use Rust at a startup, and it's been great. But I've used Rust mostly for performance-critical CLI tools, or highly specialized servers. Typically, these have a narrowly defined job that must be done reliably and efficiently. Ru
447.
by ekidd 4y ago
I've moderated a number of internet forums. And at the end of the day, you have a choice: 1. You can keep the deliberate assholes, the abusers, the white supremacists, the people with actual swastikas in their profile photos, etc. (Som
448.
by ekidd 4y ago
Several years ago, I tried some side-by-side tests of DeepL and Google Translate. The challenge was to take idiomatic humor from French websites and translate it to English. DeepL did a great job. It dealt with slang, informal grammar, and
449.
by ekidd 4y ago
Several of the byte-for-byte copies pointed out by open source authors were longer than 20 lines, and contained verbatim comments. I am not a lawyer, but that's been enough to get people in legal trouble in the US.
450.
by ekidd 4y ago
Trading systems are complex distributed systems. They need extremely high reliability and low latency, on enormous transaction volume. There are lots of complex failure modes in a system like this, and minimizing them is an advanced skill.

More ›