HackerLangs
トップ新着トレンドコメント過去質問紹介求人

iooi

no profile record

投稿

The Next Era of Knowledge Work [pdf]

cdn.openai.com
4 ポイント·投稿者 iooi·先月·0 コメント

コメント

iooi
·19 日前·議論
Nice job! I made something similar, mostly for myself: https://www.vexling.com

Plan to keep it forever free :)
iooi
·22 日前·議論
This "guy" is worth on the order of all football players put together.
iooi
·先月·議論
Ironic that the link you sent has a pretty detailed (albeit rude) answer that specifically mentions that this is an issue with surround :)
iooi
·先月·議論
> But you'd have no idea because they somehow completely deleted the neighbour's dialogue from the audio

This sounds pretty unlikely. It's more likely that there's an issue with your surround system, and that audio "should" be coming from your rear speakers but for some reason it's not.
iooi
·先月·議論
So close to being the first kilocorn. A unicorn = 1 billion, this is almost 1k.
iooi
·先月·議論
Is this because of the tok/s? Since it's pretty easy to run up a $5k bill in API usage for Claude/ChatGPT in a month.
iooi
·2 か月前·議論
I'll buy your macbook if you're trying to get rid of it!
iooi
·2 か月前·議論
The entire basis of this article is that generating tokens is a variable cost and that that cost will not decrease over time.

> On an economic basis, a monthly subscription only makes sense with relatively static costs.

Running a data center is a fixed expense. Whether or not people use that data center to it's capacity doesn't change how much the operator pays (electricity use factors into this, since a GPU running at 100% will use more watts than an idle one, but it doesn't move the needle much on other fixed and variable costs of a data center).

> They also assumed, I imagine, that the cost of tokens would come down over time, versus what actually happened — while prices for some models might have come down, newer “reasoning” models burn way more tokens, which means the cost of inference has, somehow, gotten higher over time.

This is backwards. When the cost of something goes down, people use it more. This is basic supply and demand. Inference has gotten cheaper already, and will continue to do so.

Companies subsidizing costs for growth happens all the time. Yes, switching to usage-based pricing instead of subscriptions sucks for customers, but enterprises will continue to pay.
iooi
·4 か月前·議論
Is 100 t/s the stadard for models?
iooi
·4 か月前·議論
In what way?
iooi
·7 か月前·議論
SATA, not NVMe, they will still be making SSDs.
iooi
·10 か月前·議論
Google started out by sponsoring Firefox, then hired many of their key developers to build Chrome. Cloudflare is likely doing the same thing, they know that strategically they depend too much on Google for the browser. This will get their foot in the door without making a large commitment. If the project goes well, it will be absorbed into Cloudflare in a few years.
iooi
·10 か月前·議論
Wrong. India is the most populous country, and the third highest emitter.
iooi
·7 年前·議論
OT, but I love this sentence juxtaposed with the motto for your blog: "brevity is for the weak"
iooi
·7 年前·議論
> I dislike the rationale of "CloudFlare is a private company. They can do whatever they want" ... like wut?

Doesn't CloudFlare have the same free speech rights to express their own views towards hate speech (i.e., by not providing service to white supremacists)?
iooi
·7 年前·議論
Even CP? I don't think you've thought this all the way through.
iooi
·7 年前·議論
I can highly recommend nand2tetris. It doesn't go as far as this post but it covers a lot and is a ton of fun to work through.