Algorithms. For decades, they’ve been the gatekeepers of the tech industry, the sacred rites of passage for anyone hoping to secure a coveted job at a top-tier company. Entire platforms like Leetcode and HackerRank have built empires around this obsession, turning the act of coding into a gladiatorial spectacle. But let’s be honest: is the ability to invert a binary tree under duress really the hallmark of a good programmer? Or is it an elaborate hazing ritual we’ve all agreed to endure?
The famous quote, “Why memorise what’s in books when you can look it up?”—often attributed to Albert Einstein—has long been a rallying cry against rote learning. Yet the tech industry continues to worship at the altar of algorithmic trivia. In a world where artificial intelligence can churn out optimised solutions in seconds, this fixation isn’t just outdated—it’s laughable. The truth is, Leetcode isn’t preparing you for the future; it’s clinging desperately to the past.
The Algorithmic Cult
Let’s not mince words: algorithmic interviews are elitist nonsense. They reward those with the privilege of time and resources to grind hundreds of obscure problems while penalising talented developers who dare to prioritise real-world skills over party tricks. They’re a relic of a time when programming was about eking out every ounce of performance on machines with laughably limited resources.
But here’s the truth: that time is long gone. Today’s developers rely on libraries, frameworks, and tools to solve 99% of the problems algorithms were once meant to address. Now, with AI coding assistants like GitHub Copilot, ChatGPT, and Claude, even that final 1% is falling away. Why should anyone memorise how to implement Dijkstra’s algorithm when an AI can do it for you faster, better, and with fewer off-by-one errors?
The defenders of this archaic system will argue that it’s not about memorisation—it’s about problem-solving. But let’s be real. Solving a contrived puzzle under artificial conditions has nothing to do with designing a scalable system, debugging a gnarly production issue, or navigating the political minefield of a project meeting. Algorithms are just the industry’s way of gatekeeping, of keeping out the “unworthy.”
The Leetcode Lie
Leetcode is dying, whether it wants to admit it or not. Platforms like it have thrived on the illusion that their problems are the ultimate test of programming prowess. But AI has shattered that illusion. Why waste hours practising solutions to problems you’ll never encounter in the wild when an AI can solve them and explain the solution in plain English?
This isn’t just a change in tools; it’s a change in mindset. For years, the industry has conflated technical ability with the ability to perform well in these algorithmic parlour games. But as AI continues to outperform even the most seasoned Leetcode grandmasters, it’s becoming painfully obvious that this emperor has no clothes. And the companies that cling to these outdated interview practices are just exposing their irrelevance.
Algorithms in the Age of AI
To be clear, algorithms themselves aren’t the problem. They’re tools, and like any tool, they have their place. But their place isn’t in hiring processes or the day-to-day workflow of most developers. The skill we need today isn’t memorising how to implement quicksort from scratch—it’s knowing when to use it or when to use something else entirely.
This is where AI changes the game. Tools like ChatGPT can generate algorithmic solutions and explain their trade-offs. They don’t just give you the fish; they teach you how to fish. And they do it without the elitist baggage of traditional education or the ridiculous pressure of a timed whiteboard interview.
The real skill in the AI era isn’t solving algorithms—it’s framing the problem correctly. It’s knowing which tool to use, understanding the system you’re working within, and making decisions that align with business goals. Algorithms are no longer the be-all and end-all. They’re just one small part of a much larger picture.
Goodbye, Leetcode. You Won’t Be Missed.
As AI takes over the grunt work of programming, Leetcode and its ilk will go the way of the dodo. Companies that cling to these methods will be left behind, while those that embrace the new reality will thrive. The future of hiring won’t be about who can perform the best under artificial conditions; it will be about creativity, adaptability, and the ability to collaborate with AI effectively.
This shift isn’t just inevitable—it’s necessary. For too long, the tech industry has fetishised skills that are increasingly irrelevant while undervaluing the qualities that actually make someone a great developer. AI is forcing us to re-evaluate what matters, and it’s about time.
Embrace the Revolution
The rise of AI is a wake-up call. It’s not just changing how we code but what it means to be a programmer. The obsession with algorithms, the grind of Leetcode, the absurdity of whiteboard interviews—it’s all crumbling. And good riddance.
So, who cares about algorithms? Not the future. If you’re still clinging to them, you’re not just out of touch—you’re obsolete. AI isn’t just changing the game; it’s rewriting the rules. And if you can’t keep up, don’t be surprised when the industry leaves you behind.
This argument extends to why bother to learn to play a musical instrument or write music, machines can do it a lot faster and often more perfectly. But sometimes the imperfect is what makes something perfect!
Agree why learn something when you can look it up in a book, makes sense or does it?
Imagine if medics had to look up everything and the internet was down or unavailable, would they be able to treat you. We still need an independence of knowledge and ability, relying on AI for everything might well be our downfall.
Software engineering is not so much about coding as creating solutions to problems and requirements, coding is just a tool to achieve a certain end. Machines can help to achieve that, although at the moment would prefer to view them as a very efficient way of looking up information and not always perfectly!
Think your thesis has some traction and validity, but perhaps not 100% so!
I think I have mixed feelings about your comments. I agree that focusing on algorithms during an interview is too low-level to be that useful. But it does depend on the job I guess. Jobs down at the metal demand the kind of esoteric knowledge you denigrate, along with knowledge about things like cache mechanics, memory barriers, multi-CPU programming, and so forth. For more conventional situations I think an interview that revolves around system design and so forth is much more helpful.
I have somewhat of a concern about the future of AI. Let’s not forget that it is – per the name – artificial. It is a [very] sophisticated parrot in some ways, and right now it is very useful for some things. But my future concerns rest on the manner of how AI is trained, which is that it is trained on publicly available knowledge for the most part, and right now that public knowledge is not heavily tainted with AI-generated content. How good with AI be when it is “learning” from content that was actually generated by AI in the future?
Maybe I am becoming obsolete… I’ve been a developer for nearly 40 years now. I have not written a conventional algorithm like linked lists etc for perhaps decades because eg. C++ or C# have their own implementations. What I still need to consider as a developer, however, is things like the Time Complexity of certain algorithms, and general efficiency, especially for multi-threaded situations that may endure contention. AI may also provide guidance in these areas, but the less generic the situation, the more a developer is going to have to adapt things to their own situation.