It's a good article, but the headline feels a bit off considering the source. The idea of personal responsibility has been co-opted by corporations in the past to deflect from systemic issues, so seeing it framed this way- especially given the ownership and agenda of WP- feels a little manipulative. Though, to be fair, I am almost certainly over-analyzing it.
Not at all, I think it's certainly worth pointing out. Regular people (like the author of this article, presumably) often end up inadvertently parroting the agendas that have been instilled into them, consciously or not.
Given current models can accomplish this task quite successfully and cheaply, I'd say that if/when that happens it would be a failure of the user (or the provider) for not routing the request to the smaller, cheaper model.
Similar to how it would be the failure of the user/provider if someone thought it was too expensive to order food in, but the reason they thought that was they were looking at the cost of chartering a helicopter form the restaurant to their house.
Realtime LLM generation is ~$15/million “words”. By comparison a human writer at the beginning of a career typically earns ~$50k/million words up to ~$1million/million words for experienced writers. That’s about 4-6 orders of magnitude.
Inference costs generally have many orders of magnitude to go before it approaches raw human costs & there’s always going to be innovation to keep driving down the cost of inference. This is also ignoring that humans aren’t available 24/7, have varying quality of output depending on what’s going on in their personal lives (& ignoring that digital LLMs can respond quicker than humans, reducing the time a task takes) & require more laborious editing than might be present with an LLM. Basically the hypothetical case seems unlikely to ever come to reality unless you’ve got a supercomputer AI that’s doing things no human possibly could because of the amount of data it’s operating on (at which point, it might exceed the cost but a competitive human wouldn’t exist).
They can cheat non-anonymous interviews too. An alternative is to have candidates go in person to an office to interview, but the grading and hiring panel only sees anonymized recordings of the interview.
Holy hell this looks really cool! its about 70% of what i need it to be out of the box! and it has a decent api as well. I might just end up using this. Anything you have in mind on the retrieval/ search side of things?
I think it really depends on how you want to use it. There are 'Collections' which are basically like custom categories you can create as well as #tags. When you add a bookmark you can put it directly into one of your collections and add tags. I have created a few collections around high level areas but have generally just added bookmarks without sorting, then after I read them I will move them into a bucket. The highlighting feature is very nice as well.
I would try it out for a few days and see how you like it. I definitely think there is something psychologically behind saving links and never reading them, I have quite the backlog.
> I definitely think there is something psychologically behind saving links and never reading them, I have quite the backlog.
It's like having peace of mind or hoarding—you never know when you might need those links down the road, but knowing they're stashed away somewhere is pretty comforting.
I used Raindrop with 7,000 links, and I usually go back to some of them to recall or confirm information.
There’s a guy here on HN who tried to “fix” bookmarks or start page. He also sells wallpapers for his extension. Idk how to find the post though. Maybe will update later.
This seems like an interesting way to approach this problem. obviously this is different for everyone, but do you find a chronological approach useful? also do you manually edit the webpage? if you have some code thats helping you manage your system itll be cool if you can share it. I like the approach here, and making it a webpage makes it super easy to do some of the other things i wanted as well
Reverse chronological shows the most recent first which is useful.
It's a manual system. Web pages, HTML and style are manually edited with a text editor. There is more code, HOWTOs and notes on my private home pages. Here is a public page:
Another mechanism I am looking to implement as a web extension is this: all my duckduckgo searches are also ran against a local save of my bookmarks, and get displayed on the side. atleast i think it will help me
