KODE Labs’ Post

Global Innovation at Its Finest

Potential Google breakup ⚖️, Tesla Robotaxi rumors 🚕, software engineer pay map 💰 In today's issue of TLDR... 👇 📱𝗕𝗶𝗴 𝗧𝗲𝗰𝗵 & 𝗦𝘁𝗮𝗿𝘁𝘂𝗽𝘀 DOJ indicates it's considering Google breakup following monopoly ruling (3 minute read): https://lnkd.in/eh6kugB6 Tesla Robotaxi reveal: What to expect (8 minute read): https://lnkd.in/eY82CiqD 🚀 𝗦𝗰𝗶𝗲𝗻𝗰𝗲 & 𝗙𝘂𝘁𝘂𝗿𝗶𝘀𝘁𝗶𝗰 𝗧𝗲𝗰𝗵𝗻𝗼𝗹𝗼𝗴𝘆 Google identifies low noise “phase transition” in its quantum processor (6 minute read): https://lnkd.in/eht9ZPJK Zap Energy shows off its new fusion power prototype, Century (5 minute read): https://lnkd.in/eF3rv5Va 💻 𝗣𝗿𝗼𝗴𝗿𝗮𝗺𝗺𝗶𝗻𝗴, 𝗗𝗲𝘀𝗶𝗴𝗻 & 𝗗𝗮𝘁𝗮 𝗦𝗰𝗶𝗲𝗻𝗰𝗲 Announcing Deno 2 (16 minute read): https://lnkd.in/gvjpJRXh Don't let dicts spoil your code (7 minute read): https://lnkd.in/gKG2EhDA 🎁 𝗠𝗶𝘀𝗰𝗲𝗹𝗹𝗮𝗻𝗲𝗼𝘂𝘀 Software Engineer Pay Heatmap (Website): https://lnkd.in/gvmnf_AY The Internet Archive is under attack, with a breach revealing info for 31 million accounts (2 minute read): https://lnkd.in/e-3q3Yuq ⚡𝗤𝘂𝗶𝗰𝗸 𝗟𝗶𝗻𝗸𝘀 Eating less can lead to a longer life: massive study in mice shows why (6 minute read): https://lnkd.in/gCrSMssB Amazon's new warehouses will employ 10x as many robots (2 minute read): https://lnkd.in/eB9sTYiF OpenAI projections suggest no profits until 2029 (1 minute read): https://lnkd.in/gmhWhVDe From software to reality? (2 minute read): https://lnkd.in/eZFVWksh The Open Source AI Definition RC1 is available for comments (4 minute read): https://lnkd.in/etmDZ6qF Are these 18 things all that's killing you? (7 minute read): https://lnkd.in/gHik6kg9 If you enjoyed this post... Subscribe to our free daily newsletter: https://lnkd.in/gB_buC2E

🌟 Reflecting on an Inspiring Experience at Google I/O Extended in San Francisco! 🌟 I recently had the incredible opportunity to attend the Google Developer Group (GDG) Silicon Valley I/O Extended conference in San Francisco, and it was an unforgettable experience! 🎉 The event was a treasure trove of knowledge, innovation, and networking. Much talk was about Open Source AI, AI security, and privacy. Highlights of the conference included: 💻 Gemini Code Assist & Vertex AI: Jason Smith's in-depth session on Gemini Code Assist was enlightening. The demonstration of real-time code generation and contextual understanding of code was nothing short of revolutionary. 🧠 Prompt API & Chrome Built-in AI: Hemanth HM delivered an engaging and informative talk on the Prompt API and Window.ai. His insights into AI capabilities were both fun and thought-provoking. He also showcased a fascinating experiment on browser AI-powered chat that operates without needing a network connection, highlighting the potential for offline AI applications. 💡 Code Generation with LLMs: Daniel Goncharov's session on code generation using Gemini and other LLMs was a standout, showcasing the future trajectory of AI-driven development. 🌍 Scaling Global Applications with AI: sujata tibrewala's presentation was truly inspiring. She demonstrated the profound impact of AI on scaling massive global applications, highlighting real-world applications and potential. 🚀 Prompt Engineering Workshop: Peter Danenberg's mini-workshop on prompt engineering was a fantastic hands-on experience, enriching my practical understanding of AI applications. Connecting with so many talented professionals and discussing the latest trends in AI was a highlight of the event. The energy and innovation in the room were palpable, and it was an honor to witness the incredible work being done in this field. 🤝💬 A heartfelt thank you to Sakshi Gupta for organizing such a phenomenal conference. Your dedication and effort made this event truly memorable! 🙏👏 🍕Pizza was the cherry on top! #GoogleIOExtended #SanFrancisco #AI #GeminiCodeAssist #VertexAI #PromptAPI #WindowAI #CodeGeneration #PromptEngineering #Networking #TechCommunity

When you look back at the impact of technology over the past few decades, it's clear that it will continue to be a huge part of our lives. It's obvious with the increase in its popularity that moving forward, AI will also be a large part of our lives. Now is your chance to dip your toes in both or refresh your knowledge and meet other like minded individuals. Join Google Developer Groups as they come to Merced to discuss Responsible AI. #GDGDevFest #TechFuture #MercedTech

Is this the end for software engineers? 😳 #Devin is a new model from Cognition that behaves like a real software engineer. It is an autonomous agent that can perform complete engineering tasks using its own web browser, shell and code editor.   For example, it can: ✅ Learn to use unfamiliar technologies ✅ Build and deploy applications from start to finish ✅ Autonomously find and fix bugs in codebases ✅ Train and refine its own AI models   Cognition engineers evaluated Devin on the SWE-bench, a challenging benchmark that asks agents to solve real-world GitHub issues found in open source projects. Devin correctly resolved 13.86% of the issues, outperforming the state of the art by nearly 12% 🤯

AlphaCode2 was released by Google , and in the tech report we get a fascinating view under the hood https://lnkd.in/ePT3BYxz At this stage, I think it is very fair to say that solving real world problems using LLM's is a software engineering challenge Alphacode2 has 4 Gemini instances , 1 for scoring , rest as finetuned models to get to the right answer see the hoops it has to jump through (in a simplified way, refer to the paper for more details ) 1 - 1 million samples generated 2 - of the 1 million 95% are eliminated as they dont give the right answer 3 - rest of the 50,000 samples are clustered , scored by a model to pick top 10 and it gets these right 43% of the time in manual evaluations 🙀 🤷♂️

I think GenAI can probably help with technical debt, which means we might have more time for security in terms of software engineering. Of course, it's also possible that GenAI will introduce technical debt, or make obvious mistakes like flipping a digit from positive to negative and wreak havoc, but that's what I want to find out. So I wanted to start in an unsafe but relatively known place, which means building a simple guestbook application that I can use to see if GenAI can help fix and secure it and make it better overall. Blog post - https://lnkd.in/gqbXV9B9 The next TAICO meetup is on Sep 17th. Register here - https://lnkd.in/gimesyvp

It's never too late to start learning more and more about the digital world we live in. Excited to embark on my own journey of learning about Backend Engineering and Computer Systems, inspired by friends Pratyaksh Gupta and Tinku Choudhary. I am starting my own #500DaysofBackEnd challenge, where I'll explore at least one topic a day to gain a better understanding of what builds the digital age we live in today. Day 1 of #500DaysofBackEnd starts with exploring how the internet works and various protocols that drive it. #backendengineering #computersystems #digitalage

🌐👩💻 Devin: The AI that writes code, debugs like a pro, and even trains its own models. Say hello to the future of software engineering! #DevinAI #TechInnovation #llm #devin #softwareengineer #jobloss #aiart #cognition #karankulx #3minuteai #karankulx

Get started with the most powerful LLM library: AdalFlow--The PyTorch library to build, eval, and auto-optimize any LLM tasks. Though only public for two months, AdalFlow has been widely adopted by tech companies in Silicon Valley, including LLM engineers from Meta and various top tech startups. 1️⃣ Create your first chatbot, RAG, and Agent—all in one Colab notebook https://lnkd.in/g23YDaHt 2️⃣ Deep dive on Generator https://lnkd.in/dCRhXg2U 3️⃣ LLM evaluation playbook https://lnkd.in/gTvD93mP 4️⃣ Get started with end-to-end Auto-optimization -- Question Answering: https://lnkd.in/gtECHuGi -- Classification: https://lnkd.in/g4py2e6w 👉 Github repo: https://lnkd.in/g-KN9-dV Happy building! 😊 #adalflow #artificialintelligence #machinelearning #llms