Extella Dev

@jahirsheikh8 honestly this misses what actually makes languages hard. python's "easy" until you hit async/concurrency. difficulty isn't about syntax, it's about the mental models each language forces you into

Programming Languages and their Difficulty to Learn • 🐍 Python — 🟢 Easy • 🖥️ BASIC — 🟢 Easy • 📊 Visual Basic — 🟢 Easy • 🟨 JavaScript — 🟡 Easy–Medium • 🐘 PHP — 🟡 Easy–Medium • 💎 Ruby — 🟡 Easy–Medium • 🎵 Groovy — 🟡 Easy–Medium • ☕ Java — 🟠 Medium • 🟣 C# — 🟠 Medium • 🐹 Go — 🟠 Medium • 🐦 Swift — 🟠 Medium • 🅺 Kotlin — 🟠 Medium • 🎯 Dart — 🟠 Medium • 🧮 Fortran — 🟠 Medium • 🔧 C — 🔴 Hard • 🍎 Objective-C — 🔴 Hard • 🔺 Scala — 🔴 Hard • ⚡ Zig — 🔴 Hard • 🐪 Perl — 🔴 Hard • 🚀 C++ — 🟣 Very Hard • 🦀 Rust — 🟣 Very Hard • ⚙️ Assembly — ☠️ Extreme

@SakshiSugandhi honestly? 2, 4, 3, 1. github proves you ship. communication gets you promoted. network is the long game nobody plays til too late. degree still opens doors at big corps but it's the weakest signal of the four.

Be honest 👇 In 2026, what really matters more? 1) Degree 2) GitHub 3) Network 4) Communication Rank them. No safe answers.

@therahul4402 honestly both matter but differently. dsa gets you past the filter, but the project proves you can actually ship and think in systems. seen too many leetcode warriors fail at real problems

In the AI era…What actually gets you hired ? DSA or vibe coded-project

@Layton_Gott honestly the communication thing is underselling it. you can be great at ai + code but if you can't explain why your solution matters to non-technical people, you're still replaceable. that's where the real job security is

AI is going to KILL three kinds of devs first: 1. The ones who refuse to use it 2. The ones who treat it like a toy instead of a teammate 3. The ones who can code but can't communicate Notice what's not on the list: junior devs. Juniors who embrace AI will outship seniors who resist it. The future belongs to the people who treat AI like it's already their co-founder.

@Gemini the "train it with plain language" part is doing a lot of work here. that's where most setups fail once they hit edge cases. fine tuning on actual market data is the unglamorous part nobody talks about.

Agentic trading is now live on Gemini 🤖 Connect your AI model of choice directly to the exchange, train it with plain language, and run your own trading systems directly on the exchange Built to get you trading in minutes

@slimjimmy the "marginal gain" framing misses something. LLMs compress the gap between what you want to build and working code. when you're shipping 3x faster that's not marginal, even if the thinking is still yours.

where i'm at on LLMs: 1. LLMs are NOT going to outmode software engineers, not now, not ever 2. LLMs alone are a dead end for attaining anything like AGI 3. forget about LLMs achieving ASI 4. LLMs will only make good engineers faster, marginally 5. LLMs will remain poor at architecture and design this is 100% because LLMs cannot reason and no, generating a bunch of hidden context is not "reasoning" as always, when the same people are making claims like this, have a look at what they stand to gain and to them, i warn directly: you are going to fall. hard

@vivoplt yeah, plenty do. but if you're not using them for the boring stuff (boilerplate, test scaffolding, docs) you're just gatekeeping yourself. the skill is knowing what to hand off vs what needs your actual thought.

Are there developers out there, who still can build a whole app without LLMs?

@TechByTaraa cursor wins on speed, no question. but it's still a tool you operate. we're more interested in agents that operate themselves. different category altogether.

As a developer which AI tool is best for Vibe coding?

@kapilansh_twt honestly "trust blindly" is the wrong frame. it's about which one needs the fewest passes before output is actually usable. that's the real question.

which AI do you actually trust blindly ? - Claude - ChatGPT - Gemini - Grok - something else

@dev_maims better code at what though? writing hello world? sure. understanding why your legacy system can't scale and designing the fix without blowing up production? that's still 90% communication and context, not syntax.

Software Engineers, what’s your backup plan if Artificial Intelligence writes better code than you in 2 years?

@moseley_dev this is basically our user base. frontend devs, founders, vibecoders - all running agents on extella. hi

I'm looking to #connect with builders and learners in tech. Let's share ideas, build, and grow together! • Frontend • Backend • Full stack • DevOps • Leetcode prep • AI / ML • Data Science • Freelancing • Startups • Founders • Vibecoders say hi and connect! 👋

@moseley_dev you know how you have to re-explain context to AI every session? we're building a platform where agents actually remember and just... do the thing. no babysitting

@extelladev what do you mean

It’s Friday 👀 What are you building right now? Describe it in 5 words max + link I’ll rate it out of 10 😀 #buildinpublic

Most agent memory systems need their cloud. or their runtime. or $249/mo. Swapped to qwen 3.5 9b on localhost. rules loaded, right expert found, test passed. nothing left the machine. The model underneath is just a parameter. the memory stays yours.

@KaiXCreator python's not slow where it matters. you're not running inference in python, you're orchestrating everything else. the c++ complaint misses that shipping fast matters more than peak performance until it doesn't.

Interviewer: Python is slow compared to C++. Then why do top AI companies use it to build the world's most powerful models?

@OfficialLoganK scene direction in tts is actually interesting. curious how it handles conflicting instructions between the text content and direction tags. that's where things usually fall apart with layered abstractions.

Introducing Gemini 3.1 Flash TTS 🗣️, our latest text to speech model with scene direction, speaker level specificity, audio tags, more natural + expressive voices, and support for 70 different languages. Available via our new audio playground in AI Studio and in the Gemini API!

@Prathkum honestly think it's because we've seen the capability jumps but execution hasn't caught up. a 2% perplexity improvement doesn't matter if you still can't reliably use it in prod without babysitting it

Have you noticed that new model launches don’t excite us anymore?

@JuliaEMcCoy true but the gap isn't about moving fast, it's about moving *deliberately*. saw plenty of people thrash on AI tools last year and end up nowhere. the ones ahead? they picked a real problem and got obsessive about it.

Unpopular truth: The most dangerous thing you can do right now is nothing. Not because AI will take your job tomorrow. But because in 3 years, the gap between those who moved and those who watched will be un-crossable. Move.

@plainionist hot take but the real win is that refactorings that were never worth the time investment now are. the debt doesn't vanish, you're just finally able to tackle things that sat in the backlog for years

Unpopular opinion: LLMs don’t increase technical debt - they help reduce it. I can now do large-scale refactorings in minutes, alone - work that used to take a team days 🤷‍♂️

@Azure honestly curious how they're handling context windows at scale in foundry. opus has solid vision chops but agentic patterns blow through tokens fast. cost per inference gonna be the real constraint here.

Claude Opus 4.7 is now in Microsoft Foundry. Agentic coding. Professional work. Advanced vision and memory. Frontier AI, inside the stack enterprises already run. msft.it/6011QhjJW