π AI Visionary | π Tech Explorer | π Lifelong Learner | Cybersecurity Advocate | Let's grow and secure the future of tech together! ππ‘β¨
@thsottiaux Blue is useless, much better models on Claude (Mythos for researchers), Cyber with Git integration in the cloud was nice, but now billing model is terrible for researchers. You lost your edge.
Has anyone else noticed free AI models (ChatGPT & Grok) struggling with simple tasks lately? π€
Had 3 basic requests fail recently. Is this actual quality degradation or are we just past the honeymoon phase and thinking more critically?
Share your experiences! #AI #ChatGPT #Grok
Copilot for VSCode is great, huge difference comparing to the early release. What could be improved: long sessions management, add native support for task/session including session log for better resume, native proposal to companyβs code standard, tech guide agent must follow. It is all possible to setup today manually, native support or recommendations would me it far better for enterprises
AI is writing code now? Sure, just like my intern 'writes code.' After years of building software, I can tell you: AI's not taking my job. It's making me coffee. β Here's why AI fails at real dev work (from someone who's tried everything) π§΅ #DevTruth#AILimits
8/8 What are your experiences with these "autonomous" AI agents? Have they delivered real value or just flashy demos that fall apart in practice?
Share belowβI'm genuinely curious if anyone has found them consistently reliable. π
#AIAgents#AIHype#TechTruth
π₯ Hot take: Autonomous AI agents like Manus AI are 90% marketing hype, 10% actual value.
They promise the world but deliver buggy, impractical outputs that rarely work in real life.
AI can help youβbut "autonomous" is a dangerous myth.
Prove me wrong ππ§΅
7/8 Real talk: If you're paying premium prices for "autonomous" AI agents right now, you're funding their R&D while beta testing buggy products.
The future is human-AI collaboration, not replacement.
Anyone claiming otherwise is selling something. π°