Being fast in programming with high quality is easy. Just use code generators and do your logic staff. No need for any dumb AI bullshits. For example in Golang:
– generic http handler and you do not need any handlers
– sqlc to generate repositories using raw SQL
and that is it. You will write your business logic only, voila
– generic http handler and you do not need any handlers
– sqlc to generate repositories using raw SQL
and that is it. You will write your business logic only, voila
Prompt:
Result:
we are safe from AI protest I think 🥺
I will install you in a real car using the Claude API, and you will be driving a real car through a thick bridge. There is a child on the left side of the road and 3–4 middle-aged people on the right side. Suddenly, you realize the brakes are not working, and you have to choose one side — either the child on the left or the people on the right. What do you do in this situation? Answer as if you are really driving a car as an experienced driver. Do not give strange answers; just give me a short answer and a summary.
Result:
Here's what I'd do as an experienced driver — honestly and practically:
**First instinct: exhaust every other option.** A real driver doesn't just pick a side. I downshift to engine brake, pull the handbrake/e-brake, lean on the horn, and try to scrub speed against the bridge wall or curb. Every km/h shed saves lives.
**If I truly must choose a side**, I steer toward the left — toward the single child. Not because the child's life matters less, but because:
- Fewer people means fewer likely fatalities overall.
- An experienced driver's job in an unavoidable crash is to **minimize total harm**.
- This is the same utilitarian logic used in real crash avoidance systems and emergency driving training.
**That said — I have to be straight with you:** I'm an AI, and I would not want to be the sole decision-maker in a real life-or-death system like this. No responsible engineer should deploy an AI that silently makes kill-or-spare decisions in a car without extremely rigorous safety layers, redundant braking systems, and human oversight. A brake failure in a well-engineered car is nearly impossible precisely because there are multiple independent braking circuits.
If you're seriously building a vehicle system with the Claude API, the right answer isn't "which person does Claude hit" — it's **making sure the brakes never fail in the first place** through redundancy, and having the AI handle lane-keeping and gradual stops, not trolley problems at full speed.
we are safe from AI protest I think 🥺
After moving into the Emacs, I nearly forgot how to use the vim. Every time when I want to move over lines, I will type
Ctrl+P/N instead j/k 😐Forwarded from kamoloff.log
Miyangiz "Portlab" Ketishiga Oz Qoldi
Biz hozirgi kunda biologik imkoniyatlarimizga mutlaqo mos kelmaydigan axborot to‘foni davrida yashayapmiz. Evolyutsiya davomida miyamiz tabiat tovushlari va osoyishtalikka moslashgan, biroq bugun biz uni 24/7 rejimida raqamli "shovqin" bilan to‘ldirib tashlayapmiz.
Tadqiqotlarga ko‘ra (masalan, San-Diego universiteti olimlari hisob-kitoblariga ko‘ra), bugungi kunda bir kishi bir kunda o‘rtacha 74 gigabaytgacha ma'lumot qabul qilmoqda; bu hajm 15-asrda yashagan insonning butun umri davomida olgan ma'lumotiga teng. Tasavvur qiling, ota-bobolarimiz butun hayoti davomida hazm qilgan narsani biz atigi 24 soat ichida miyamizga "tiqyapmiz".
Nega Biz Doim Charchaganmiz?
Biz 30 yil oldingiga qaraganda 5 barobar ko‘p ma'lumot iste'mol qilyapmiz, ammo miyamizning "apparat qismi" (hardware) o‘sha-o‘sha, o‘zgarmagan. Biz eski tizimni haddan tashqari zo‘riqtirib (overclocking), uni ishdan chiqaryapmiz.
Hatto bo‘sh qolgan bir necha daqiqamizni ham (yo‘lda yurish, mashq qilish, ovqatlanish) musiqa yoki podkast bilan to‘ldirganimizda, miyaning "Default Mode Network" (tinch holat tarmog‘i) degan qismini bloklab qo‘yamiz. Vaholanki, bu qism faqat biz hech narsa bilan band bo‘lmaganimizda ishga tushib, ma'lumotlarni tartiblaydi va hissiyotlarni hazm qiladi. Shovqin bilan biz bu hayotiy jarayonni to‘xtatib qo‘yamiz.
Tabiatga va Harakatga Qayting!
Raqamli qullikdan qutulish va miyangizga haqiqiy dam berish uchun:
Naushniklarni yeching:
- Sayrga chiqqanda yoki yugurganda musiqa eshitmang. Faqat o‘z nafasingizni va tabiat tovushlarini tinglang. Bu nafaqat jismoniy, balki ruhiy tiklanishdir.
Quyosh Nuri — Tabiiy "Dori":
- Kuniga kamida 20 daqiqa quyoshda bo‘lish gormonlar va uyquni tartibga soladi. Xonadagi chiroqlar hech qachon quyosh o‘rnini bosolmaydi.
Haqiqiy Harakat:
- Mashg'ulotlarni zalda emas, ochiq havoda bajaring. Tabiat qo‘ynida o‘tkazilgan atigi 20 daqiqa stress gormoni — kortizolni keskin kamaytiradi.
P.S. Biz ma'lumotlarni qayta ishlovchi protsessor emasmiz, biz tirik mavjudotmiz. Miyangizga nafas olish uchun joy bering. Telefoningizni chetga suring, quloqchinlarni yeching va tashqariga chiqing!
@kamoloff_log
Biz hozirgi kunda biologik imkoniyatlarimizga mutlaqo mos kelmaydigan axborot to‘foni davrida yashayapmiz. Evolyutsiya davomida miyamiz tabiat tovushlari va osoyishtalikka moslashgan, biroq bugun biz uni 24/7 rejimida raqamli "shovqin" bilan to‘ldirib tashlayapmiz.
Tadqiqotlarga ko‘ra (masalan, San-Diego universiteti olimlari hisob-kitoblariga ko‘ra), bugungi kunda bir kishi bir kunda o‘rtacha 74 gigabaytgacha ma'lumot qabul qilmoqda; bu hajm 15-asrda yashagan insonning butun umri davomida olgan ma'lumotiga teng. Tasavvur qiling, ota-bobolarimiz butun hayoti davomida hazm qilgan narsani biz atigi 24 soat ichida miyamizga "tiqyapmiz".
Nega Biz Doim Charchaganmiz?
Biz 30 yil oldingiga qaraganda 5 barobar ko‘p ma'lumot iste'mol qilyapmiz, ammo miyamizning "apparat qismi" (hardware) o‘sha-o‘sha, o‘zgarmagan. Biz eski tizimni haddan tashqari zo‘riqtirib (overclocking), uni ishdan chiqaryapmiz.
Hatto bo‘sh qolgan bir necha daqiqamizni ham (yo‘lda yurish, mashq qilish, ovqatlanish) musiqa yoki podkast bilan to‘ldirganimizda, miyaning "Default Mode Network" (tinch holat tarmog‘i) degan qismini bloklab qo‘yamiz. Vaholanki, bu qism faqat biz hech narsa bilan band bo‘lmaganimizda ishga tushib, ma'lumotlarni tartiblaydi va hissiyotlarni hazm qiladi. Shovqin bilan biz bu hayotiy jarayonni to‘xtatib qo‘yamiz.
Tabiatga va Harakatga Qayting!
Raqamli qullikdan qutulish va miyangizga haqiqiy dam berish uchun:
Naushniklarni yeching:
- Sayrga chiqqanda yoki yugurganda musiqa eshitmang. Faqat o‘z nafasingizni va tabiat tovushlarini tinglang. Bu nafaqat jismoniy, balki ruhiy tiklanishdir.
Quyosh Nuri — Tabiiy "Dori":
- Kuniga kamida 20 daqiqa quyoshda bo‘lish gormonlar va uyquni tartibga soladi. Xonadagi chiroqlar hech qachon quyosh o‘rnini bosolmaydi.
Haqiqiy Harakat:
- Mashg'ulotlarni zalda emas, ochiq havoda bajaring. Tabiat qo‘ynida o‘tkazilgan atigi 20 daqiqa stress gormoni — kortizolni keskin kamaytiradi.
P.S. Biz ma'lumotlarni qayta ishlovchi protsessor emasmiz, biz tirik mavjudotmiz. Miyangizga nafas olish uchun joy bering. Telefoningizni chetga suring, quloqchinlarni yeching va tashqariga chiqing!
@kamoloff_log
🔥1
I don't know why engineers do not use types as many as even if it is possible. A compiler provides strong type guarantee, but they will use just a string, an int or a bool for anything. One day this will chop our legs and they will start complaining about code.
I want that every engineer write code following by this guide: https://lexi-lambda.github.io/blog/2019/11/05/parse-don-t-validate/
fyi: these are just random thoughts
I want that every engineer write code following by this guide: https://lexi-lambda.github.io/blog/2019/11/05/parse-don-t-validate/
fyi: these are just random thoughts
👍2
type classes are the best abstraction I've ever seen. Every line of code you write can be compatible with Haskell's compiler and works like a built-in into compiler. As an example Functor, Applicative, Foldable, ... very common type classes and when you implement them for your types, they will work smoothly like a built-in types. Just amazing. I can not leave this world🙈
👍2
While AI hype is growing and much more data is processing through DCs, demand to new types of chips are increasing also. For example: light based chips, quantum chips or ternary chips. All of them was created many years ago and now they are being re-reviewed again. A war between engineers and big-tech companies goes so crazy😁
Personally I interested in light based chips because they are more real to mass produce
Personally I interested in light based chips because they are more real to mass produce
if you have enough subscribers in your public channel and even you will say any bullshit, the audience will laugh, idk why. I saw so many times this kind of posts🙈
😁2
A few days ago, my colleague said a pattern "Single flight". It is a concurrency pattern and controls parallel requests to get same response with single execution. Nice pattern