rco8786

23 ms·

511.
by rco8786 2y ago
Interesting. I consider myself a rubyist and never considered this. Perhaps because the rest of the language is so concise that this little verbosity never really bothered me
512.
by rco8786 2y ago
> This should have been obvious before the fact to anyone who understands how CSS selectors work in browsers. For sure, but that's somewhat reductive. This is not exactly common knowledge. Certainly not something you would expect an
513.
by rco8786 2y ago
There’s another option here though. Human supervised tasks. There’s a whole classification of tasks where a human can look at a body of work and determine whether it’s correct or not in far less time than it would take for them to produce t
514.
by rco8786 2y ago
I think there’s a long way to go also. I think people expected that AI would eventually be like a “point and shoot” where you would tell it to go do some complicated task, or sillier yet, take over someone’s entire job. More realistically i
515.
by rco8786 2y ago
> Very close to FAANG senior salary Base salary maybe. But more like ~40% of FAANG TC. (Which only furthers your point)
516.
by rco8786 2y ago
> The biggest problem with Rails is that people get so glued to the conventions that everything gets shoehorned into a model or controller instead of using a better abstraction. As a Rails fanboy, I couldn't agree more. Rails'
517.
by rco8786 2y ago
Thanks so much! Any resources for deploying something like this into a cloud/datacenter and standing up an API?
518.
Ask HN: How to host an AI model that processes PDFs into structured data?
6 points by rco8786 2y ago|7 comments
519.
by rco8786 2y ago
> We found that getting to accuracy that is reliable enough for automation is challenging. This is in the problem description of your pitch, and leads me to believe that tile.run has been solving this problem. Is that right? > Coming
520.
by rco8786 2y ago
It's a fairly "natural" case for AI, and there is tons and tons of people who need to pull structured data of out of PDFs for myriad reasons.
521.
by rco8786 2y ago
Can you speak to the accuracy, particularly of numerical value extraction, that you’re achieving? I have a use case for pulling tabular financial data out of PDFs and accuracy is our main concern with using AI for that type of task.
522.
by rco8786 2y ago
After using Cursor (another AI focused fork) I'm 100% on the fork train. AI built natively into the IDE presents another layer of speed and isn't subject to the limitations of the extension system (which is awesome in its own righ
523.
by rco8786 2y ago
Is there a comparison with Cursor I can read?
524.
by rco8786 2y ago
That's been gone since Rails 4 or 5 (we're on 8 now)
525.
by rco8786 2y ago
I think historically that's probably right, but something that has gotten much better with time. Previously, ActiveRecord gave newer programmers a legible, fluent interface to the database in a language they already knew. So instead of
526.
by rco8786 2y ago
https://github.com/ruby/rbs is the "officially supported" one. It's terrible, as is Sorbet. No offense to anyone involved with either of the projects, they just miss the mark in pretty fundamental ways.
527.
by rco8786 2y ago
> It's the job of a manager to know what their reports are up to, and whether they're doing a good job of it, and are generally effective. I don't think this is mutually exclusive with building tools that gives a bird&#x27
528.
by rco8786 2y ago
Yes, I think agree. And when you use a calculator and get a result that doesn't make sense to you, you step in as the human and try and figure out what went wrong. With the calculator, it's typically a human error that causes the
529.
by rco8786 2y ago
I mean sure, but this is exactly the argument against using a calculator and doing all your math by hand also.
530.
by rco8786 2y ago
Have you actually used Cursor? It does not seem like it. I'd suggest trying it before dismissing it. > Copilot seems to only be aware of code in current file Cursor does not have this limitation
531.
by rco8786 2y ago
The founders did an episode on Lex Fridman's podcast that covers the technical sides of Cursor far better than I ever could: https://www.youtube.com/watch?v=oFfVt3S51T4 But to summarize, Copilot is an okay ChatGPT wrap
532.
by rco8786 2y ago
Curious why you don't just use Cursor's built-in chat?
533.
by rco8786 2y ago
> I think AI is still nowhere near where it needs to be to provide business value in software development. And you go on to say that your experiment was to build a TODO app "some time ago" in a single file of code.
534.
by rco8786 2y ago
There's a lot of people in these comments who are talking about the state of off-the-shelf LLMs for writing code, and they are missing the point of this article. The article is about Cursor, the IDE. It is much, much more than a ChatGP
535.
by rco8786 2y ago
> But does great at generic stuff, which is already trivial to find on stack overflow. The major difference is that with Cursor you just hit "tab", and that thing is done. Vs breaking focus to open up a browser, searching SO, f
536.
by rco8786 2y ago
Before you dismiss all of this because "You could do it by hand just as easily", you should actually try using Cursor. It only takes a few minutes to setup. I'm only 2 weeks in but it's basically impossible for me to ima
537.
by rco8786 2y ago
It’s not stopping them. It’s just causing them to change tactics. NAR still has a stranglehold even after the lawsuits.
538.
by rco8786 2y ago
To be fair to realtors. There is A TON more work and expertise that goes into any transaction than just taking pictures and listing the property.
539.
by rco8786 2y ago
To be fair, NAR goes to great lengths to maintain their monopoly and lock out anyone who tries to buy or sell a house without using them....specifically because they know they're priced insanely high for the services they provide and a
540.
by rco8786 2y ago
IF they can deliver a product that actually works, yes.

More ›