another_twist

Its useful for answering questions related related to back references. eg who was this character if you're reading a novel with too many of these.

I don't use LLM for writing just factual research stuff. And this would happen even in those questions.

Same here, had to configure ChatGPT to stop making these statements. Also had to configure bunch of other stuff to make it bland when answering questions.

+1. My favourite bit is when pivots are chosen randomly in quicksort, we get linearithmic expected complexity. The CLRS proof using indicator random vars was a oh-shit moment.

Its weird how object detection models are "AI" now. These models and their weird errors have been around for quite a while. The issue is vendors claiming that there is no chance of errors. Ideally you would have a 2 eyes system such that if AI has a tolerable false positive rate and have a human review. But of course, you cant fire people with AI so why would we do the sane thing.

And of course there are policy wonks who would make gun ownership a human rights issue even though its fundamentally unsafe to have such free gun ownership.

Definitely one of the As in the FAANG. In fact both the As have terrible recruiting practices. One is a known ghoster and given that you were ghosted after a senior level meeting tells me which one.

If this is the worst this guy has done, well done Canada !

The author mentions simplicity in their Readme. I would be very interested to read their journey and some of the decisions they made where they preferred simplicity. More of this please !

Vibe coding doesnt mean the author doesnt understand their code. Its likely that they don't want carpal tunnel from typing out trivial code and hence offload that labor to a machine.

Never ran into this. This seems odd, I use the 2025 build and it works just fine for me. With maven and gradle.

I use it for Java. I have never used anything else and never had any performance issues on my 16GB MacBook Air.

If its Webstorm maybe its because of automatic refresh capability ? I've had perf issues with VSCode as well with autobuild enabled. Autocomplete would grind to a halt.

TIL uta has a brilliant website. And apparently no tuition, well done them.

Merge and Quicksort are good ones too. Quicksort esp since the idea has a flavour of "I could come up with that". Its a beautiful algorithm.

When the array is almost sorted. Bubble sort complexity is linear + inversions so if the inversions are low (the more sorted the array the lower the number of inversions), bubble sort is close to a linear pass.

I do codeforces in my spare time. Sometimes I implement and ML paper. Other times, I like to slog through my implementation of Raft, Paxos and VR. Not everybody wants to build generic crud app number 1,200,674. Coding is for solving problems, the problems might be engineering or just pure fun.

Not really, its quite easy to tell if you havent prepped well. The place where AI is good is online assessments.

This is a good product, the mechanism for me was an excel sheet. I wont sign up though, I've ground enough LC. These days I dont even prep for algorithm rounds and still manage to land offers. But I'd have appreciated this when grinding myself.

Used to be in the same camp here until I had to interview for a specialist role. I'd happily swap Leetcode rounds and doing away with the highly subjective - design a class hierarchy nonsense.

> I remember a few decades ago somebody saying the JVM was incredible technology, and as a user and programmer I still have zero clue what the hell they could have been thinking was good about the JVM.

I see what you mean. In that case we can add Scala backed systems as well to the JVM balance sheet. If we simply look at the JVM and the systems it backs, there's very little evidence that it isnt a marvel of technology. It powers more impactful systems than few other technologies.

> The global financial system moves so much money around that comparisons to GDP are a bit silly.

Agreed. I guess its comparing production capacity to distribution capacity. Distribution capacity will equal n_tx * tx_amt. Having said that, another metric to look at is how much of software infrastructure is built on Java. Simply adding AWS to this equation proves the value added by Java backed systems. Hard to say that about any other langauge. Also we can look at versatility, Java is used to write very large data processing systems, CDN networks, API servers and even widely used consumer apps (IntelliJ products). Its very hard to find any other language that has had an outsized impact across domains. Of course the counter being Linux written on C powers all of the internet. True but C doesnt have the cross domain impact that Java has had.

So I disagree with the assessment that Java is a terrible langauge performance or productivity wise or it wouldnt have had this impact.

Good riddance: I guess Java is on a purge right now. First public static void main, now this unused mess. Its good for the ecosystem.

Java running in the browser is unlikely as typescript has largely tamed the mess of Javascript. Java requires a JVM and shipping an entire JVM so its runs atop another VM is kinda redundant. Except if JVM itself gets compiled and cached as a WASM bundle and Java compilers start accept WASM-JVM as a target. That will just be distraction tbh, Java has its strength in large scale systems and it should just focus on those rather than get caught up in Frontend's messy world.

I understand the sarcasm but this take is devoid of fact. Modern Java loads fast, Java 21 has pretty good functional programming featurez. The ecosystem churns out language level features at a pace and a budget that would put most large funded startups to shame.

Java is also the workhorse of the big data ecosystem and moves enough money either as product revenue or as transactions than most nations GDP. They didn't figure out startup times for 10+ years, they were busy dealing with Oracle and its messy management. I think it will simply continue to get better given that Java has endured through so many language fads. It has its ways to go but it will end up like SQL - here before we were alive and will be here when most of us are dead.

> My mother pointed out that a lot of the songs along the lines of “my own true love proved false to me” were about unplanned pregnancies.

I was this years old when I realized it.

Its also a math problem. The kind I've encountered that make bad decisions are also the ones shockingly bad at doing back of the envelope calculations.

Honestly failing candidates in an interview put of a sense of superiority is just about saddest thing I've heard. I mean how lonely do you have to be ?

/endrant.

Case in point - PyTorch vs Tensorflow. The Pytorch team and Soumith in particular was everywhere. You could ask about it in the forums, Twitter, freaking reddit and there would be an answer.

The risk of tech debt is marginal cost of adding features goes up as tech debt goes unpaid.

Say this in an interview and its a perfect way to fail, even though its true. Its sad how interviewers often take pleasure in pointing out that anything said outside their packets is a signal for lack of technical knowledge. I've been in and passed several tech interviews. I've also interviewed plenty of people, if someone points out the human aspect of a problem, I actually award points. Sad how often I have to fight with my colleagues.

"But what about using a message queue.."

"Candidate did not use microservices.."

"Lacks knowledge of graph databases.." (you know, because I took a training last week ergo it must be the solution).

No it isnt. Next word prediction is what humans do to communicate anyway so the criticism isnt valid. Except you do that for your own sentences (if you do it for others its considered rude :) ).

Anyways this criticism is now dated given that modern day LLMs can solve unseen reasoning problems such as those found in the IMO.

It does have something to do with the stock market, since its about making hypotheses and trading based off that. However, I'd agree that making a proper trading AI here would require reasoning based fine tuning for stock market trading actions. Sort of like running GRPO taking market feedback as the reward. the article simply cant do that due to not having access to the underlying model weight.

True that and you left Succession off the list :).

Thats why I think its a good addition. Its a step up from their usual quality. They are anyways paying 72B for the company. Netflix has distribution and Warner Bros delivers quality. I think its a perfect fit. I mean imagine Warner getting acquired by Amazon and folded into prime video - that thing where search returns episode names rather than the series. That would be like Rolexes being sold at the local Walmart. There's only two homes for quality content and thats Apple TV and Netflix. Thats why I think its a good deal.