Theo @itheo051

China is so ahead of the game👏

You Called Us Xenophobic Not Long Ago,FACTSSSS!!!!!

Dr.Dennis Adison Ouma. @DrDennisOuma

7 days ago

-KICK ILLEGAL FOREIGNERS OUT OF. KENYA, IMMEDIATELY 🙂🙂🙂 - Africa is not one country. -Kenya belongs to Kenyans. -South Africa belongs to South Africans. -Ghana belongs to Ghanaians. -If you are in Kenya illegally,leave -Get back legally. -Sovereignty is not xenophobia.

563 625 2K 760K 144

🚨 93-year-old Dr Katrina Esau is the LAST fluent speaker of South Africa’s ancient 25,000-year-old N|uu language. Based in Upington with her granddaughter Claudia Snyman, she needs AI developers & linguists to digitize her 45 unique click sounds into a living voice model before it’s too late. Let’s connect tech giants with her school to safeguard this indigenous treasure we can train AI to keep this language. Please RT! 🔄 South Africa 🇿🇦 everything matters. @GaytonMcK @Harvard @OpenAI @GoogleOrg

🔥 AI Engineer Roadmap - 2026 Stage 1: 🐍 Python Stage 2: 🔧 Git + GitHub Stage 3: 🗄️ SQL + Databases Stage 4: 📊 Machine Learning Fundamentals Stage 5: 🧠 Deep Learning - PyTorch Stage 6: 🤖 LLM Fundamentals Stage 7: ✍️ Prompt Engineering Stage 8: 📚 RAG - Retrieval-Augmented Generation Stage 9: 🧩 Vector Databases + Embeddings Stage 10: 🛠️ AI Agents - Tools, MCP & Function Calling Stage 11: 🚀 AI Deployment - Docker + Cloud Stage 12: 💻 Build & Ship AI Applications 🏆 Complete these + build real projects = AI Engineer unlocked.

a lot of people ask me how to get started in robotics, and my answer is always simple: start building. robotics is one of those fields where you can spend years preparing to begin. you can study calculus, linear algebra, control theory, embedded systems, computer vision, mechanics, electronics, ROS, reinforcement learning. all of these matter. but you don’t need to understand all of them before touching a robot. pick something embarrassingly small and make it move. build a two wheel robot. make a servo track an object. build a tiny robotic arm. balance an inverted pendulum. make a drone hold altitude. your first project will immediately expose the gaps in your knowledge. suddenly PWM isn’t an abstract concept because your motor needs it. PID isn’t another equation because your robot is oscillating. coordinate transforms matter because your arm needs to reach a point. filtering matters because your sensors are noisy. mathematics becomes useful because reality gives you a reason to learn it. this is the learning loop i recommend for robotics: build → fail → identify the missing concept → study it → implement it → fail differently → repeat. don’t try to finish the robotics curriculum before becoming a roboticist. let the machine construct the curriculum for you. robotics is too interdisciplinary to learn linearly. build something slightly beyond your current ability and allow every failure to tell you what to learn next.

#InCaseYouMissedIt

keyboard_arrow_left Previous keyboard_arrow_right Next

Miss Ziprusha @destinyzee

2 weeks ago

#InCaseYouMissedIt Good Morning 🙌

keyboard_arrow_left Previous keyboard_arrow_right Next

128 155 1K 141K 73

@destinyzee WOD: Violence

Traditional DE vs AI Data Engineer👇

I spent 8 hours learning dbt, so you don't have to I spent a workday reading the dbt docs, the Snowflake performance guides, and the BigQuery partitioning reference to find out where the minutes actually go. The answer was uncomfortable, because it isn’t a dbt problem you can fix with a dbt flag. dbt is a compiler and a scheduler. It renders your Jinja into plain SQL text, works out an order, opens a few connections, and sends the text down the wire. Then it sits there. That reframe matters because it changes what you go looking for. You stop tuning dbt and start reading query plans So a slow project is one of four things, and all four are warehouse-side: ✅ You picked the wrong materialisation for the model’s size. ✅ Your incremental strategy is scanning far more than it needs to. ✅ Your DAG shape forces the same work to happen more than once. ✅ Your tests are quietly running hundreds of full-table queries. Read the complete newsletter here -blog.datavidhya.com/p/i-spent-8-ho…

#InCaseYouMissedIt

keyboard_arrow_left Previous keyboard_arrow_right Next

DEAR SOUTH AFRICANS, THE GIRLS DID IT ONCE AGAIN. WE’RE INTO THE LAST EIGHT 🔥🔥🔥🔥🔥🔥🔥🔥🔥🔥🥳🥳🥳🥳🥳🥳🥳🥳🥳🥳🥳🥳🥳🥳🥳🥳🥳🥳🥳🥳🥳🥳🥳🥳🥳🥳🥳🥳🥳🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦 WELL DONE GIRLS, THE COUNTRY IS PROUD OF YOUR ACHIEVEMENT @Banyana_Banyana 🔥🔥🔥🔥🔥🔥🔥🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦🇿🇦 #WAFCON2026

keyboard_arrow_left Previous keyboard_arrow_right Next

Most data engineers skip this concept (very important for interviews) If you run your pipeline twice and the numbers change, it's broken. Compare the output tables before and after. 👉🏻 If row counts doubled → your load is append-only. 👉🏻 If totals change 👉🏻 If it crashed on a primary key This property has a name: idempotency. Same input, same run, same result, no matter how many times you execute it. 3 changes that get you 90% of the way: ✅ Stop INSERT-ing into tables. Overwrite the partition you're processing, or MERGE on a business key. ✅ Use the scheduler's execution date as a parameter. ✅ Make the run own a clear slice of data: one date, one partition, one chunk. Junior engineers build pipelines that work. Senior engineers build pipelines that can be rerun at 2 AM by someone who's never seen the code.

@SuperSportTV Everybody busy comparing RSA and Nigeria, while Australia is crushing everyone and everything. 🙆

@groldrush What if we burn them 1st!!!!

@TrackGazette Being number one is good, but betting Canada is always better.

@Kiminino72 Is it Korra's fault too??

@Eleven__11XI @Kiminino72 Explain.

Dear Africans We won. We are the best. We are South Africa.

@PhutiSemenya14 Go to Qwaqwa and you'll win by a landslide.

@DearS_o_n Proverbs 24:16 NIV [16] for though the righteous fall seven times, they rise again, but the wicked stumble when calamity strikes.