🚀 Built a full DEX from scratch on ICP, with efficient liquidity, robust swap & LP mechanisms, transparent fees & scalable multi-canister architecture 🔧
Always solving complex blockchain challenges 💡
@StarkNetPro @ICPdevelopers@kirat_tw#DeFi#Rust#ICP#Crypto#Web3
I am really loving the impact reading books is creating on my work. I am now more consistent and more focused thanks to "Deep Work" book which taught me how to do productive work in less time.
Now I am able to pay more attention and focus at what I am building
#buildinpublic
40% done with Andrew Ng’s ML course.
Skipping the hype and actually learning the fundamentals.
If your feed is tired of "AI wrappers" and wants real engineering, let’s connect.
What are you building today? 🛠️
#BuildInPublic#DataScience
I believe that if there were only 1 or 2 language in blochchain and more interoperablity between them then we could have been advanced more than now as it would have been more developer friendly and we could have spent time more on logic and #engineering
One this I like in #MachineLearning that it's mostly based out in #python unlike #blockchain where you may have to switch to multiple languages for multiple chains.
It really decrease the burden to switch to completely different language every time
#engineering#AI
@ash_twtz Dude, why are you so pessimistic about gemini? Its really powerful amd good for many task and even better for some tasks than the ones you mentioned
NumPy makes our lives much easier. But is relying on it good for someone doing core #AI/ML research?
I’m wondering if understanding things from scratch matters more when working on model efficiency or simpler approaches—or if abstractions like NumPy are enough.
#machinelearning
Please welcome GPT-6 Sol and GPT-6 Luna to the GPT-6 universe.
GPT-6 Sol and Luna build on the advances behind GPT-6 Astra, bringing much of its strengths into faster and more affordable models to support work at scale.
We’ve also made caching and inference more efficient, and we’re passing the savings directly to you: 50% lower API prices for Sol and Luna compared with GPT‑5.6 promotional pricing.
Maximize your #ML research with free GPU resources.Students can bypass hardware limitations using these platforms:
• Google Colab: Access to NVIDIA T4 GPUs in-browser.
• Kaggle: 30 hours of free weekly GPU compute.
• Azure : $100 in cloud credits.
#MachineLearning#DataScience
@rasbt Do the MiMo-V2.6 results suggest that scaling and improving the RL/data pipeline is now a bigger source of capability gains than designing new attention architectures? What do their ablations show?
@AndrewYNg Agree completely, @AndrewYNg
.
The more I learn the fundamentals of ML, the less “scary” AI feels. Most fear comes from misunderstanding, not from the actual technology. Glad you’re pushing back on the hype