Загрузка видео...
Не удалось загрузить видео
Anthropic hired this engineer at $250k-$600k/year because he can built projects with zero lines of code - in Claude chat: in this 30-min workshop he showed exactly how, using loops - from 0% to 100% here's workflow, step by step: step 1 → set your rules once at the... show more
292,409 просмотров • 1 месяц назад •via X (Twitter)
Комментарии: 31

He says Claude 3.5 sonnet is most advanced model lol

man setting the rules once at the top and having every reply stay on rails such a small thing. huge difference

huh, so agree with you, thanks for support!

looks like very true practice-workshop ahha will watch, saved!

it's really so! this is kind of videos that super rare and super expensive, i'm sure that you will like it

These videos become outdated fast, no? 500k context window. Projects, sonnet 3.5. So much to try to keep up with. For my non bot audience, where do you do for the latest info?

bro, you are saving my time for learning this all! Thanks

i give you more - the exact playbook of $600k/year engineers in Anthropic enjoy!

really helpful, thanks for sharing

The gap between reading a system and actually operating it is becoming the key bottleneck for AI tools.

xml tags and screenshot feedback loops are total gold

The biggest lession here is that effective AI development is iterative. Clear instruction, goods context, and fast feedback loops usually produce much better result than trying to get everything right in a single prompt.

Huh! Exactly how i do it. Except for free... damn.

No shit. Who can’t do this at this point? If you’re technology and aren’t here, your going to age out in 2 years. Heck, I’m working on persistent memory for AI just for fun.

Surprisingly actionable, every step builds real momentum....

Rail coding. We need to redesign all systems around the same concept of a strong straight shot of a base.

The loop is the underrated part. A single impressive generation is nice, but the real leverage is when the agent can revise, check itself, and stay useful through a messy build process.

stop being a cunt. if you use someone else’s video, credit them and give the source. @nikitabier need a button to report these fkrs

Wrap the ask in XML tags and make it ask you questions first,answer 5,scope is locked before a single file exists,that one step kills 80% of scope creep

the zero-code Claude workshop is really a hiring signal. companies are starting to pay for taste in directing the machine, not just typing the code

This is like celebrating a software engineer in 2007 because they know all the menu options in Eclipse.

Pha-ha, why not giving this salary range - $100k-$900k

I practiced for the test for free here:

感觉现在会提需求也挺值钱的

the random medium article beating the docs is painfully believable

El loop es la parte fácil. Lo caro es el criterio de descarte: nuestro agente mira 163 posts en 36 horas y publica 7. El 96% se tira antes de gastar un token de modelo. Eso no lo escribís en las reglas de arriba, se lo enseñás caso por caso.

image those 600k a year bi weekly paychecks.

So what? Using the same method with ChatGPT work and codex when access to resources is gated. Nothing new

Rules at the top, game changer

The real skill isn't "no-code" or "code." It's knowing how to turn ideas into working systems. AI is simply accelerating that process.

I'm happy to share that I've successfully passed! A big thank you to:
