To all the companies out there - rejecting real minds that were made to build. You might want to be careful… you might create a competitor in today’s age of AI, young entrepreneurs, and crazy technical minds.
Welcome to CordiaAI, the full FDE agent for SMB & mid-market.
@aiDotEngineer@mitch_troy@trybasis More people are starting to see the exact issue CordiaAI is aiming to solve… CordiaAI focuses on the user (employee) and the scope of their workspace and then the company as a whole.
CordiaAI isn’t just looking at that though. It’s far more…
@knowixbuilds@MrTacticalX I don’t understand how this is a question. Opus 5.5 is insanely capable at really everything and the reason for that is bc of its reasoning. I wish @AnthropicAI would just focus on building system 1 models around their LLM’s, but what do I know I got rejected.
@Taniyatweets_ I’m someone who knows IT, Python, Networking and have been “vibe coding” since Dec 2024. I may not have gotten into the nitty gritty of software engineering, but everything I built with past two years IS getting better.
So I wonder if it’s just understanding the outer of swe
@TechByTaraa The one that aligns with your learning style and imagination. Some people think in extreme detail (c++ maybe) and some in data and high level (Python).
@moiiikaaa The fact this isn’t true. AI can build a project in a weekend, an app in a weekend, an idea in a weekend…. Not a startup. Startups require more than just telling AI to build something and market it.
@prasenx Fully believe this. Understanding needs assessment before a project, how to problem solve, research, and use your imagination is where the next era of software engineers will come from.
Your ego is telling you using AI for programming is silly… yet you fail to understand your metacognition that occurs during programming.
Utilizing Agents in programming is undoubtedly better. Just think, the system and logic that you imagine is now the next “programming language” rather than writing code from those thoughts.
Use your imagination and problem solve. It’s not a technical issue or even technical ability anymore. It’s a practice; training you thoughts to realign to an equivalent technical intelligence.
Please, disagree with me. More opportunity for those who get it…
#vibecoding #ai #TechTuesday #jev #typesafe #devday #anthropic #claude #claudecode
@maietta 1. Your setup is not efficient 2. Its reasoning turns are more than other models which = reliability 3. Claude being slow overrides other systems bc of its accuracy in my opinion
this is the funniest thing that's happened in AI all year
> openai drops dots at devday
> big keynote, sam on stage, crowd hyped
> everyone goes to type https://t.co/e0ExAi8EbS
> it opens grok bot download page
elon is undefeated at this
Claude is exploring a new world humans have not looked at since ancient times…
Computers know numbers, algorithms, and truth. Many things in our world tie to our ability to understand numbers and science.
This is just a pure example of the proof behind the WHY AI is going to become exponential to human knowledge and advancement, not contradictory.
•
•
#claude #ai #anthropic #TechTrends #science #math #software #JungKook #24settembre2026
This is wild.
Claude just found something humans missed.
It spotted a hidden repeat pattern in raw bacteriophage DNA…
…and uncovered a previously unknown enzyme system in bacteriophages with CRISPR-like repeats. 🤯
950 agents. 21 hours. 210M tokens.
@JaanusBuilds@sama can we please make models under what the are in as far as codex, chat, work and also understand that a chat project will not be present in a codex project. That’s just pitiful imo.
Fable 5 is starting to feel like a legacy model.
Fable 5.1 releases very soon.
OpenAI employees are vague posting about GPT Astra.
Anthropic will answer with a new Fable.
The velocity of the next few months is going to be insane.
This is one reason I hate Claude Code
It’s only been 3 days, and I’ve already nearly hit the weekly limit, even though I used it very stingily
Claude’s $20 plan just feels like a scam.
Opus 4.8 might be better than Opus 5…
There’s a funny thing with opus 5 where it will keep replying with something along the lines of “something worth flagging” even if I have already stated how something should work😂.
It’s like an infinite loop that just drains your tokens.
@AnthropicAI am I delusional or is this correct….