Just to drive the point home, in 2019 the total VC market was ~$300 billion. To date, roughly $235 billion is tied up in just OpenAI ($168b) and Anthropic ($67b)

    - music apps and gapless playback: basically all major car manufacturer software is trash, I'd challenge you to find a car software that's better than Tesla
    - stretches without superchargers: to be clear, not stretches without places you can charge the car, right? Just places where Tesla's fast charging stations exist. Tesla has an order of magnitude more of these type of stations than any other similar network
    - Service centers: I have no knowledge here. From how silly the other arguments are, I kind of assume that this isn't fully true, but I'll give you the benefit of the doubt and say it's true. Not awesome, definitely an area they need to get better in. But lets not forget, major manufacturer service centers are not exactly know for being quick/good/cheap
    - Color options: I guess it would be cool to have truly custom colors, but basically every mainstream car ever has had a limited set of colors to choose from, and individual dealerships almost never have every color of that limited set available

I really like the idea of a side by side comparison. I'll build that out next!

Best part is the url updates in real time, so you can copy the url (or "Share this offer" button) to show someone else what an offer represents.

- If you're a recruiter, this is an easy tool to help a candidate understand your offer

- If you're asking a friend/mentor/spouse for advice, this is an easy way to give them context

Hope it's helpful!

Comments URL: https://news.ycombinator.com/item?id=45405319

Points: 2

# Comments: 3

If one task changes the button to green, and the other one changes it to red at the same time, something will have to decide how to reconcile that difference

Fun story, strong recommend

Their overall rate of accidents is estimated as 30 times higher than conventional elevators

Germany saw an average of one death per year due to paternosters

Small nit: the initial rollout of Gemini was a dumpster fire. It is now quite good. For my use cases, I can't get any other current LLM to give me better results than Gemini 2.0 Flash. It's also free

But even that kind of proves your point, right? Pretend you 100% believe me without verifying. That means in a year or two the winner has transferred between ~3 companies. This is not a cheap mantle to keep passing around. The AI wars are going to get heated over the next year or two

It doesn't have any pressures to make money, in fact it's basically a free place for PR and advertising jobs for portfolio companies

I think because of this, a bunch of companies/tools have tried to hop in this space and promised the world, but often times people are best served by just hitting OpenAI/GPT directly, and jiggling the results until they get what they want. If you're not comfortable doing that, there are even companies that do that for you, so you can just focus on the prompt itself.

So that being said, help me understand why I should be adding this whole system/process to my workflow, versus just hitting OpenAI/Anthropic/Google directly?

Too many times, when people complain about javascript, what they are really complaining about is the hectic nature of web development. You have three trillion dollar companies (Google/Chrome, Microsoft/Edge, Apple/Safari) actively fighting over agreed upon standards for how the dom is rendered and manipulated. Try and learn any language in that environment, and you'll have a very frustrating time

Why functional programming? Javascript was one of the earlier mainstream languages to support passing around functions as first class citizens, and it's often an aspect of the language that gets downplayed. By and large, javascript sits at the sweet spot of allowing you to do cool functional programming concepts, without all the strictness and verbosity of a language like Haskell. Is that a good thing? It depends, obviously. But for the sake of "Change my mind", this is where I bet you'll start looking at it in a different light

I'm not saying it's the best/one-true language. It's just another cool tool to have on your tool belt

Make a web crawler that hits publicly traded tech companies careers page, once a day, and tracks how often their job listings change. Make a big line chart at the top of a landing page that compares the different companies job openings (by count), over time. Penalize listings that list/delete/re-list job openings to always make it seem like they are hiring. Maybe have a max time a job post can be open before it's no longer part of their count (3 months?)

What you'll end up showing is similar to this article: A lot of company's stock is evaluated on growth, and part of the growth estimation is based on how much they are hiring in this down market. Some companies know this, and are trying to game the system

Anyways, I bet a decent amount of people would watch that like a hawk, and the honest companies would love it because it would show how great they really are doing

gpt-4, gpt-4-turbo, and gpt-4o are not the same models. They are mostly close enough when you have a human in the loop, and loose constraints. But if you are building systems off of the (already fragile) prompt based output, you will have to go through a very manual process of tuning your prompts to get the same/similar output out of the new model. It will break in weird ways that makes you feel like you are trying to nail Jello to a tree

There are software tools/services that help with this, and a ton more that merely promise to, but most of the tooling around LLMs these days gives the illusion of a reliable tool rather than results of one. It's the early days of the gold rush still, and every one wants to be seen as one of the first

$12, fantastic documentation and tutorials, and you can literally just plug it in via usb and edit the python code as if it's just a text file on a thumb drive. No programmers, special IDE's, or specialty equipment

Microcontrollers are fun again

This is likely @pierat's point: sure, it's possible, but your diy version will be objectively worse in a variety of ways while also taking you a pretty long time to put together.

    - when your function calls don't have enough white space between them, they'll some times mix up their return value (crosstalk)
    - the more deeply your if/else statements are nested, the more random their results end up being (voltage/IR drop)
    - when the user makes a selection, your bound function is called somewhere between once and a dozen times. Each time with a different value (bounce)
    - their is no `true` or `false`. You just are given a float between one and zero, where one is true and zero is false. Sometimes `true` is as low as 0.3 and sometimes `false` is as high as 0.7
    - the farther apart your variable declaration is from it's use, the less it represents what you actually declared. If it's two lines apart, a 'foo' string is still 'foo'. Hundred lines apart though, and 'foo' might become 5.00035 (attenuation)