Cyclone_

I've heard of plenty of CTOs coding at small startups, but the surprising part is he has no direct reports.

I am assuming that when they say AI they just mean LLMs and not other cases of deep learning. Even so I would agree the $20 billion seems low.

I've used it for creating a template for an app. It usually creates a pretty accurate app with at least a basic UI and some interactions. The biggest issue is that once you've created the template, it's pretty tough to add some more complicated features using the AI.

I liked the search results a lot better when the ads were on the side. You could see a competitor if that was of interest to you, but you didn't have to scroll when you wanted to just find what would likely be a top result.

They do provide numbers you can call if you're addicted to gambling on the ads shown on ESPN. But it feels like that doesn't do very much help.

Err I don't think that's exactly true to say there's no such thing as the best engineers. There's very few who would be capable of solving problems that someone like a Jeff Dean is able to do. Realistically though, not every company needs to solve problems of the difficulty that Google has.

The first thing that came to mind when I heard hardware we own was vehicles like a Rivian where they do run a lot of software. I can understand why they'd not want people to run software in order to avoid bad press. If someone writes something and things go wrong, it will look bad for the manufacturer, even if they're not at fault.

People leaving to do their own startups would be a pretty small portion of the job market, even on the SF Bay area.

This seems to make a good case for the brave browser

Agree that they can't retaliate through their military, but if they did it would likely be through terror attacks on civilians.

That's what Google does now for some of its AI search results. There's a link icon next to some of the answers.

In this scenario, if AI does a not so good job, there will still be good developers left to code.

I use LLMs for generating small chunks of code (less than 150 lines) but I am of the opinion that you should always understand what generated cide is doing. I take time go read through it and make sure it makes sense before I actually run it. I've found that for smaller chunks of code it's usually pretty accurate on the first try. Occasionally it can't figure it out all all, even with trying to massage the prompt to be more descriptive.

It's been about 15 years since I've used Firefox as my main browser but I'm not sure what their value proposition is at this point. They're not faster than the best browsers, and if they're not going to be better on things like privacy I don't see where they are going to position themselves. I mostly use brave now and some chrome when the occasional site doesn't work in brave.

Growing up in Minnesota I would see them quite a bit. We have a lot of lakes, so it's an environment where you would see a lot of them.

Seems like some LLM could do this. If not a chart, at least some summary of the results.

"Elegance is not a real metric".

Code maintainability certainly is something to consider that will bite you if you're not thinking about it.

Even after a 555 number was provided they still kept responding without questioning it. Are at least some of their responses automated too? Kind of funny to think of 2 LLMs negotiating with each other.

I usually wonder if they do that as well. In some cases it may be hard to depending on how data is stored. In vertica, a database I worked with would never truly delete data on disk.

I'd argue you can usually find problems on project euler that are a little more obscure than what you'd typically get on leetcode.

Well the title is a little silly even if it's meant to be provocative. There's a limit to just how many people would benefit from the drug, since those who exercise frequently enough wouldn't have the need for it ever.

I'm getting the sense there's a heavy sampling bias towards larger companies. I looked at my company which is on 1 metro area and it isn't super accurate. The titles they list don't even match up with what we call them.

I had never heard of it until now, but it looks similar to parts of the Marshall Islands in that there are very narrow strips of land.

The second paragraph of your post seems to imply negligence on the end of Uber if they're not vetting properly.

Aren't most AIs trained on puic data, i.e. this doesn't seem terribly surprising?

Agreed, scraping is more appropriate for when one gathers data from a 3rd party site.

Not justifying what they did with qc, but qc is missing from quite a few places in software development that I've been apart of. People might get the impression from the article that every software project is well tested, whereas in my experience most are rushed out.

I'd imagine at that level they'll have access to altitude tents to try and adjust. But I do wonder how effective those would be, since you only sleep in them.

From my experience it's only been electrify america chargers that have been down. I've never experienced an issue with Rivian charging, and I use that more frequently.

I'd say deep learning is a subset of machine learning, and LLMs are a subset of deep learning.