Hold on, we aren't even good at estimating but now we know how much time we saved by vibe coding? I can't wait to read the source of this info when you share it.
No they aren’t. I’ve been doing this for a few decades working everywhere from small startups to the most successful tech companies, and none of them are good at estimating.
If you read the blog post, they were able to manually migrate 3% of the tests in 1 week. Extrapolating that gives you an estimate of less than half of the 1.5 year estimate.
I’d also say there’s a pretty good chance that the 3% of tests the automated process couldn’t handle were more complicated than average and the devs would have gotten faster at doing these migrations after the first week.
It’s very unclear what was actually being compared. A team did a POC in 2023 and then some work in 2024 and then spent 6 weeks tweaking the pipeline and a final week manually migrating the tests that the automated process couldn’t handle.
But they don’t specify how many people worked on this or how many people were going to work on the original project. It could have been 1 guy on the team.
As for looking at what the actual work was, it was migrating from one test framework to another one. I’d be surprised if a team couldn’t have written a compiler to do something similar in a similar amount of time.
When you reach "The cliff of Irrational Arguments" you need to stop and reevaluate what is the purpose of the discussion.
There was a joke about the man who was threatening to jump off a cliff. No professional could convince him otherwise with sound arguments. It took a another mental patient to make an irrational threat, cut the cliff down, to scare him off.
> At first, I fought back manually, feeding them fake data. But that got old fast. So I deployed my secret weapon: a zip bomb.
> When their bot accessed my site, I served it a tiny compressed file. Their server eagerly downloaded and decompressed it, only to unleash several gigabytes of chaos. Boom. Game over.
How did you know their bot would decompress it? I thought a bot would copy the HTML content of your article, maybe the images, and paste them on their own website. At no point does it involve editing or decompressing files?
Impressive animation, by the way—the number of bots is staggering.
As soon as bots reach a page with the compressed payload, they never make another request. That's how I know it worked. Also curl, wget, or most libraries automatically decompress gzip content.
Of course, Some bots just post spam without ever reading the content back, which defeats my scheme.
It's pretty impressive that we have self-driving cars that takes us this far with Just cameras. But it probably won't get any better than this without a revamp.
That said, what if we make these self driving tools fully available to drivers. Drivers are pretty good at driving already with with relatively short training. Give them the superpowers of teslas like 360 view, predictive object path, and object detection and you have super drivers.
I don’t know why everyone says “just cameras.” Are people imagining they are using some algorithm to detect some object and come up with a distance measure? I’m imagining they are just having the cameras autofocus and determining the focal distance which seems about the same as banking on a good lidar reflection to me.
Consider how long it takes your camera to autofocus, and how sometimes it gets it wrong. Then consider how fast your car moves. In the time it takes your camera to autofocus, how far do you think a car can travel, and what happens to all the blurry things in its path that it can't see?
A camera that is slow to autofocus is the result of market compromises to reach a price segment in a platform that is within the confines of a camera body dimension. That is not reflective of the state of the art however.
Still, the whole rationale behind the camera-only approach is that cameras are cheaply sources, and therefore more suitable for mass production vehicles, unlike LiDAR which are more expensive and not as widespread.
If you're now shifting to the idea we need to use state-of-the art sophisticated camera technologies instead of the commodity stuff, now you're back to paying LiDAR prices, so why not just use that?
And anyway, Teslas sold today and for a long time now are supposed to have been sold with cameras sufficient to solve full self driving (not beta). If state of the art cameras are needed, I've got bad news for all those Tesla customers.
Hot take: you suck at CSS because you never bothered learning it.
A long time ago, everyone in my team kept making excuses why they hate css. I went to Lynda.com and found a pretty good class. I can't remember the instructor, but it was so good that I still use the same patterns more than a decade later. I tried to get the whole team to take it, but no one wanted to. "It's a waste of time", "It's not even a programming language."
They built all kinds of tooling around css, trying to avoid css. We had dormant css that no one could ever figure out if it was used, we had important and position absolute everywhere. Today, it's not so different. You see divs with 20 or 30 classes in them.
Just learn css. Any class is better than no class.
I'm glad he is back to making approachable projects. It was a boon for me during the pandemic. I particularly love the fake windows to light my closet that looks like daylight.
It's one of the ones that's been at the top of my list, but my lack of engineering/tinkering experience has made me a bit apprehensive. Is it result really worth the effort?
My friend and I looked into that one to see if it would help his mood when working in his downstairs office (his apartment had a weird layout upstairs). It kind of looks like you need to have a room you can sacrifice to hold all of the equipment, and a door with the right shape. It wasn’t very practical in that apartment, or in any apartment either of us have lived in since.
I had this insight that I thought was evident. "AGI is an interface problem". It made sense in the context. But when I decided to write about it, my initial idea and example were pretty weak. The more I wrote the more I found counterpoints. In fact, writing forced me to think through those arguments and see if the idea can still stand.
Many times, my ideas fade into darkness after a good writing session. But that's how you form a strong opinion.
reply