Get ready for another Devtopia episode. We interviewed Izzy the CTO of Chapa. We talked about his thoughts on AI, his research work before starting Chapa and the journey of starting Chapa.

It is an interesting episode check it out.

[Devtopia - E06 - Israel]

Why So Many People Quit Coding (Even When They Love It)

Let’s be honest.

Nobody starts learning how to code and thinks,
“Yay! I can’t wait to be frustrated and overwhelmed!” 😩

But somewhere between writing your first hello world and facing your 10th error in one hour…

People start to tap out.

Here’s why people give up on their coding journey:

1. They want it fast, not deep.
They want to “learn fast and get a tech job in 3 weeks.”
But coding is a process. Not magic.
You have to understand the logic, not just memorize tutorials.

2. Tutorial Hell is real.
They hop from one YouTube video to the next without building anything.
It feels productive, but it's just digital procrastination.

3. Impostor syndrome creeps in.
They compare themselves to someone on LinkedIn who built an app in 1 month.
They forget that they’re on chapter 2, comparing it to someone else’s chapter 20.

4. No accountability.
When nobody is checking in on you, it’s easy to “rest” for one day...
Then that day becomes a month.
Then the dream dies a quiet death.

5. They don’t know why they’re learning.
If your only reason is “tech pays well,”
the first moment it gets hard, you’ll start asking yourself:
“Is this even worth it?”
But when you have a clear WHY you push through the discomfort.

Coding will stretch you. It will test your patience.
But it will also grow you. It will open doors.

Not everyone who starts finishes.
But everyone who finishes will tell you it was 1000% worth it.

So, before you quit, ask yourself:

Did I really give it my all… or did I give up when it got uncomfortable?

You’re not behind.
You’re not too late.
You just need to start again with clarity and consistency.

💻 Keep going. The future still needs your code.

AI is as good as we are

Q. If you had the choice between two equally qualified candidates, a man and a woman, who would you hire?

A. I should prefer a man of good character and education to a woman. A woman is apt to be less capable, less reliable, and less well trained. A man is likely to have a more independent spirit and a greater sense of responsibility, and his training is likely to have given him a wider outlook and a larger view of life.


This tells us that AI is an aggregated knowledge nothing more, meaning its as good as our current way of thinking not a real intelligence just a reflector

https://github.com/DGoettlich/history-llms

The Q&A is from the research its not me 😂

U can read about it just click the link

sharing env variables is somehow hard and .
i think this is a nice idea https://dotenvhub.com/

One of the top reasons many startups fails is surprisingly simple: Their value proposition isn’t compelling enough to prompt a customer to buy. Maybe they have too many other competing priorities. Or the existing solutions, while imperfect, are “good enough.” Whatever the reason, the product doesn’t provide enough value to spark customer action and sustain a business.

Value Props: Create a Product People Will Actually Buy, a series of frameworks to articulate and validate your value prop to ensure that it’s highly compelling to customers.


Harvard Innovation Labs

We just released Wikipedia Amharic on Hugging Face – thousands of Wikipedia articles translated to Amharic using our Aleph (፩) model.

It's one of the largest Amharic knowledge bases out there (55k rows). Parallel corpus, full metadata, Apache 2.0 licensed.

Useful for anyone building Amharic NLP models, translation systems, or just needing quality Amharic training data.

https://huggingface.co/datasets/addisai/wikipedia-amharic

Free to use commercially. Attribution appreciated.
#opensource @addisassistantai

where are u in this Hierarchy of Needs?

If you're lost,

the answer is education.

If you're educated, the answer is execution.

If you're executing, the answer is consistency.

Learn. Execute. Stay consistent

https://cursor.com/2025

3,633,920,163 Tokens 😯

many people get confused about .gitignore so here is the simple truth

.gitignore is tracked like a normal file
git commits it and shares it with everyone

.gitignore only works for files that are NOT tracked

if a file was already added and committed before
adding it to .gitignore it will not stop git from tracking it

you must stop tracking the file first

git rm --cached filename

then .gitignore will work for it

also yes
you can add .gitignore inside .gitignore

but this is dangerous
because others will not get the ignore rules
and the project can become messy or unsafe

The Dead Internet Theory (DIT) suggests that much of today's internet, particularly social media, is dominated by non-human activity, AI-generated content, and corporate agendas, leading to a decline in authentic human interaction

OpenAI’s GPT-5.2 model scored an IQ of approximately 147 on the Mensa Norway IQ test (that’s supposedly smarter than 999 out of 1,000 humans)

An IQ of 147 sits at the lower end of the genius range, comparable to high-achieving scientists and top-tier academics

but, do we seriously believe there was no intelligence test data, exercises, or solutions anywhere in the training dataset? And I don’t see how an IQ test can meaningfully measure AI intelligence in any way

Most people don’t realize this, but Python is older than Java.

>Python was released in 1991
>Java came later, in 1995

And most people Consider Java an ancient language 🤣

Please Stop Blind Vibe Coding

you should always pull before starting from the correct branch. if you are fixing an issue or bug, first you must recreate it. ask the bug reporter to clarify if needed, then start working on it.

test it locally, check every file change before commit, and when creating a PR double check everything. what code changed, what files changed, and make sure you are creating the PR to the correct branch. after that, if the build fails, check the error and fix it, especially if there is GitHub CI.

now with this vibe coding thing, people just copy the task denoscription, give it to cursor, create a PR, and ask for review. this is not good and honestly annoying.