LeetCodin

One person on Reddit shared the following thoughts on the future of software development and software engineering in the age of AI:

Thoughts and ideas about AI killing software development:

We are building systems that are easier to start but harder to sustain.
The future of software engineering, therefore, lies not in the "end of coding," but in the evolution of complexity management.

The engineers of tomorrow will not be "vibe coders" who don't know how to code, but system architects who manage the entropy of the AI's output.

Here is the logical proof:

In computational complexity (P vs. NP), we value problems where finding a solution is hard but verifying it is easy (e.g., Sudoku, factoring).

AI coding is the opposite ("The AI Inverse-NP problem"):
- Generation is Trivial
- Verification is Hard

We are building "high entropy" systems where the cost of verification exceeds the cost of creation. We use AI to cope with that high entropy to a certain degree. Nevertheless, the complexity increases (I would argue exponentially) with the growing amount of problems being solved.

As the overall complexity (of a system) remains, software engineers will be in higher demand than ever! But the job will be different.

Some interesting reads:
- Fred Brooks wrote "No Silver Bullet, in 1986, where he is distinguishing between "accidental complexity" (syntax, compilation) and "essential complexity" (logic, state, requirements). He argued that tools could only solve accidental complexity; essential complexity is irreducible.

- Law of Conservation of Complexity: Tesler’s Law suggests that every application has a "core of complexity" dictated by the problem domain it serves. A tax preparation application, for instance, cannot be simpler than the tax code it implements. The central question in system design, therefore, is not "how do we remove complexity?" but "who handles the complexity?

- Kolmogorov complexity is the length of the shortest computer program that can produce a given string or piece of data.

I'm curious what everyone thinks.

- Do you agree with this perspective?

- How do you see the role of software developers and engineers evolving over the next 5–10 years?

- Will AI fundamentally change the profession, or will it simply shift the skills that are most valuable?

🇺🇿 Javokhir Sindarov Nomzodlar turnirida g‘olib bo‘ldi!

🇺🇸 Javokhir Sindarov won the Candidates Tournament!

1700

The full video is here: From Uzbekistan to Big Tech

To’liq videoga havola: From Uzbekistan to Big Tech

Salom!

🇺🇸
IT Park has released a teaser video where our compatriots who live in the US and work as IT specialists in large companies talk about their journey. Stay tuned - it will be interesting.

Link From Uzbekistan to Big Tech

🇺🇿
IT Park AQShning yirik kompaniyalarida IT mutaxassisi sifatida ishlayotgan vatandoshlarimiz o‘z hayotlari haqida so‘zlab bergan video tizerini chiqardi. Kuzatib boring - qiziqarli bo‘ladi.

Havola From Uzbekistan to Big Tech

Scaling PostgreSQL to power 800 million ChatGPT users

https://openai.com/index/scaling-postgresql/

Maqola juda qiziqarli chiqibdi☝️

Hammaga salom.

Bugun Google’ning Kaggle platformasida multi-agent workflow workshop ustida ishladim. Unda agentlar yaratib, har birining vazifasini tayinlab, ularga bitta kattaroq vazifa berish mumkin ekan.

Ushbu workshop’da ikki agent yaratdim: bittasiga research qilish, ikkinchisiga esa summary qilish vazifasini berdim. So‘ng asosiy run qiluvchi tool’ga savol berdim:

Savol: “If we read all the books of Nassim Taleb, what are the main ideas he wants us to really hear, written in-between the lines?”


Javob: "<h1>Nassim Taleb's Core Ideas: Beyond the Obvious</h1>

Nassim Taleb's extensive body of work, when read between the lines, emphasizes a profound distrust of predictability and a deep understanding of the inherent uncertainty that governs our world. His core message revolves around the following key ideas:

* The Pervasiveness of \"Black Swans\": Taleb highlights that history is disproportionately shaped by rare, high-impact, and unpredictable events, termed \"Black Swans.\" We tend to rationalize these events in hindsight, failing to grasp their true random nature and our profound unpreparedness for them. He critiques traditional fields like economics for their over-reliance on predictable models that ignore these outliers.

* Embracing \"Antifragility\": Beyond mere resilience (withstanding shocks), Taleb champions \"antifragility\" – the ability of systems to not just survive but actually *thrive* and improve from volatility, disorder, and randomness. He suggests that instead of seeking stability, we should aim to build systems that benefit from chaos.

* Skepticism Towards Predictions and Experts: Taleb urges a healthy skepticism regarding forecasts and the pronouncements of experts. He posits that negative knowledge (understanding what *doesn't* work) is more reliable than positive knowledge (claiming to know what *does* work), implying that true learning often comes from identifying and eliminating errors.

* The \"Barbell Strategy\": To navigate uncertainty, Taleb advocates for a \"barbell strategy.\" This involves extreme caution in most areas while simultaneously embracing high-risk, high-reward ventures. The goal is to minimize downside risk and leave oneself open to the potential benefits of unpredictable positive events.

In essence, Taleb challenges us to fundamentally rethink our relationship with risk, prediction, and uncertainty, encouraging a mindset that is not just resilient but actively benefits from the unpredictable nature of existence."

Chakki emas 🧐

🇺🇿 LeetCode'da masalalarni yechmay yurganimga 3-4 oy bo'ldi. Yaqinda yana sekin-asta qayta boshladim. Ko'p narsalarni unutganman deb o'ylagandim. Ammo bugun o'rta darajadagi (medium level) backtracking algoritmiga asoslangan masalani 10 daqiqada, birinchi urinishda yechishga muvaffaq bo'ldim. Vaziyat men o'ylagandek yomon emas ekan! 😁 Juda xursandman.

Mana o'sha masala: 3723. Maximize Sum of Squares of Digits

🇺🇸 It's been 3-4 months since I last solved problems on LeetCode. Recently, I started picking it up again slowly. At first, I thought I'd forgotten a lot. But today, I managed to solve a medium-level backtracking algorithm problem in 10 minutes on my first try. Turns out, things aren't as bad as I thought! 😁 I'm thrilled.

Here's the problem: 3723. Maximize Sum of Squares of Digits

😊👌

🇺🇸1600 LeetCode problems solved.

🇺🇿1600ta LeetCode masala -> done.

@LeetCodin

“The voice inside that's never satisfied with your achievements isn't your insecurity.

It's the part of you that knows you can do more.”
Alex Hormozi (c)

@Leetcodin

Maktab davrida eng yaxshi ko‘rgan ustozim — matematika o‘qituvchim — bir gapni tez-tez aytar edilar: «Медленно поспешайте».
Tarjimasi: “Sekin shoshilinglar”.

@Leetcodin

The hardest part of the interview is being able to relax.

@Leetcodin

Shu blogning muallifi haqida yana bir maqola yozishibdi 👉
https://the-tech.kz/kak-ustroitsya-v-deloitte-opyt-behzoda-tair/ 😊

I am now in Google Warsaw.

Hammaga salom! Vaqti-vaqti bilan obunachilar savol berib turishadi. Shunday bir savol-javobni shu yerga joylashtiraman.


Savol: "Hozirda har kuni DSA o‘rganib, LeetCode’da masalalar yechib boryapman.
Lekin hozir FAANG’ning holati unchalik yaxshi emasga o‘xshaydi, chunki layoffs bo‘lib turibdi.

Shu sababli, ishxonada o‘zimni ko‘rsatish va tezroq raise olish uchun qanday maslahat bera olasiz?"


Javob: "DSA va LeedCode bilan shug'ullanayotganingiz juda ham zo'r! Keep it up!

FAANG'da har doim layoff bo'ladi va odamlarni ishga olishni to'xtatmaydi hech qachon. Odatda "bottom 5%" ishdan bo'shatishadi.

Savolingizga 2ta maslahatim bor:
1. Boshqa tajribali dasturchilarga ham huddi shunday savol yozib ko'ring. Javoblarni yig'ib, tahlil qilib, o'zingizga to'g'irlab, keyin harakat qiling.

2. Mening "2 cents" bu haqida quyidagicha:
O'zingizga ko'proq mas'uliyat oling.
Masalan:
- Team lead'ga borib, "sizda loyihamiz bo'yicha qanday taraf (part/volume of work) borki - shuni boshqa odam qilsa, asosiy ishga qo'l urardim [asl long-term ishlarga vaqt ko'payardi]" - deb so'rab ko'ring. Loyihaning shu tarafini owner'i bo'ling. Shunday yaxshi o'rganingki - hamma sizga kelib savol so'rasin shu haqida.

FAANG kompaniyalarida siz agar mid-level'dan senior'ga o'tmoqchi bo'lsangiz, siz senior developer'ning ishini bir oz vaqt davomida qilib kelayotgan bo'lishingiz kerak.

Bu degani sizga senior'likni boshqa odam berishi kerak emas, har bir harakatingiz "Men seniorman. Senior dev bunday holatda qanday harakat qilishi lozim?" - degan fikr bilan boshlanishi kerak.

O'zingizga boshqa odamning ish o'rnidan qarab ko'ring. Masalan:
- Tasavvur qiling siz shu biznesning egasisiz. Sizga har doim persistent and accountable odamlar yetishmaydi. Unday odamlarning o'zi kamdan-kam. Agar accountable odamni topib olsangiz, uni ishda lavozimda ko'tarib, maksimal darajada qoldirgingiz keladi - to'g'rimi? Bundan xulosa - men shunday odam bo'lishim kerak."


Siz esa nima deb o'ylaysiz?
@LeetCodin

🇬🇧1500 LeetCode problems solved.

Each one a lesson, a challenge, and a step forward.
The journey continues.


🇺🇿1500 ta LeetCode masalasi yechildi.
Sayohat davom etadi.

@LeetCodin

LeetCode Rewind 2024

@LeetCodin

Jahongir (a.k.a. KaHoGi) kursdoshim o'z ORM - Object Relational Mapping yaratish kursini yozibdi.

Agar qiziq pet project qilaman va uni portfolio'ga qo'shaman desangiz, bu kurs siz uchun!