• ​​OpenAssistant released the best open-source Chat AI (https://medium.com/@sann3/open-source-conversational-ai-takes-a-big-leap-forward-e2ccc68dfb1d)
• I replaced my assistant by a bot powered by AI to do her job (https://twitter.com/nico_jeannen/status/1647396475989811201)
• UC Berkeley's open-source Vicuna LLM chatbot released new improved model weights (https://huggingface.co/lmsys/vicuna-13b-delta-v1.1)
• You Can’t Regulate What You Don’t Understand (https://www.oreilly.com/content/you-cant-regulate-what-you-dont-understand-2/)
• 91% of ML Models Degrade in Time (https://www.nannyml.com/blog/91-of-ml-perfomance-degrade-in-time)
• An open source AI tool to animate children's drawings (https://github.com/facebookresearch/AnimatedDrawings)
• Sam Altman: OpenAI is not training GPT-5 and "won't for some time" (https://twitter.com/smokeawayyy/status/1646670920214536193)
• Companies like Uber and Amazon use AI to pay people different wages for the same work, a new study finds (https://www.businessinsider.in/policy/economy/news/companies-like-uber-and-amazon-use-ai-to-pay-people-different-wages-for-the-same-work-a-new-study-finds/articleshow/99471054.cms)
• AWS announced New Tools for Building with Generative AI on AWS (https://aws.amazon.com/blogs/machine-learning/announcing-new-tools-for-building-with-generative-ai-on-aws/)
• Amazon CodeWhisperer, Free for Individual Use, is Now Generally Available (https://aws.amazon.com/blogs/aws/amazon-codewhisperer-free-for-individual-use-is-now-generally-available/)
• Current Best Practices for Training LLMs from Scratch (https://uploads-ssl.webflow.com/5ac6b7f2924c656f2b13a88c/6435aabdc0a041194b243eef_Current%20Best%20Practices%20for%20Training%20LLMs%20from%20Scratch%20-%20Final.pdf)
• AI suggested 40,000 new possible chemical weapons in just six hours (https://www.theverge.com/2022/3/17/22983197/ai-new-possible-chemical-weapons-generative-models-vx)
• BharatGPT: The Made in India AI Language Model (https://bharatgpt.live/bharatgpt-the-made-in-india-ai-language-model/)
• What Are Transformer Models and How Do They Work? (https://txt.cohere.ai/what-are-transformer-models/)
• GPT Unicorn: A Daily Exploration of GPT-4's Image Generation Capabilities (https://adamkdean.co.uk/posts/gpt-unicorn-a-daily-exploration-of-gpt-4s-image-generation-capabilities)

• RedPajama, a project to create leading open-source models, starts by reproducing LLaMA training dataset of over 1.2 trillion tokens (https://www.together.xyz/blog/redpajama)
• ChatGPT Is Consuming a Staggering Amount of Water (https://futurism.com/the-byte/chatgpt-ai-water-consumption)
• Google & UC Berkeley’s ‘Self-Debugging’ Framework Teaches LLMs to Debug Their Own Code (https://syncedreview.com/2023/04/17/google-uc-berkeleys-self-debugging-framework-teaches-llms-to-debug-their-own-code/)
• Stability AI makes its Stable Diffusion models available on Amazon’s new Bedrock service (https://stability.ai/blog/stability-ai-makes-its-stable-diffusion-models-available-on-amazons-new-bedrock-service)
• DuckDuckGo’s building AI-generated answers into its search engine (https://www.theverge.com/2023/3/8/23629095/duckduckgo-ai-answers-browser-app-extension-search)
• Major Photography Prize Winner Reveals Image Is AI-Generated, Rejects Award (https://www.vice.com/en/article/dy3vxy/sony-world-photography-awards-ai-generated)
• Community is the future of AI (https://stackoverflow.blog/2023/04/17/community-is-the-future-of-ai/)
• Soundini: Sound-Guided Diffusion for Natural Video Editing (https://kuai-lab.github.io/soundini-gallery/)
• how do we overcome 4096 token limit in OpenAI GPT requests? (https://twitter.com/wooing0306/status/1645092115914063872)
• DINOv2: A Self-supervised Vision Transformer Model (https://dinov2.metademolab.com/)
• Analyze cloud resources using GPT (https://github.com/selefra/selefra)
• GPTest VS Code Extension (https://github.com/JRB-y/gptest)
• AI-agents that automatically generate and use Langchain Tools and ChatGPT plugins (https://github.com/hey-pal/toolkit-ai)
• Drag & drop UI to build your customized LLM flow using LangchainJS (https://github.com/FlowiseAI/Flowise)

• Microsoft reportedly working on its own AI chips that may rival Nvidia’s (https://www.theverge.com/2023/4/18/23687912/microsoft-athena-ai-chips-nvidia)
• Meet the $10,000 Nvidia chip powering the race for A.I. (https://www.cnbc.com/2023/02/23/nvidias-a100-is-the-10000-chip-powering-the-race-for-ai-.html)
• Kent Beck on ChatGPT (https://twitter.com/kentbeck/status/1648413998025707520)
• Can open-source LLMs detect bugs in C++ code? (https://twitter.com/MrCatid/status/1647155706548695040)
• GitHub Copilot AI pair programmer: Asset or Liability? (https://arxiv.org/abs/2206.15331)
• Emergent Abilities of Large Language Models (https://arxiv.org/abs/2206.07682)
• AI Playground (https://play.vercel.ai/)
• Live Conversations with AI using ChatGPT and WebRTC (https://blog.livekit.io/meet-kitt/)
• T-cell vaccine for COVID-19 may last longer than current vaccines (https://www.psu.edu/news/research/story/t-cell-vaccine-covid-19-may-last-longer-current-vaccines/)
• If there's one thing AI has done for us, it's killed Zuck's metaverse (https://www.pcgamer.com/if-theres-one-thing-ai-has-done-for-us-its-killed-zucks-metaverse/)
• An example of LLM prompting for programming (https://martinfowler.com/articles/2023-chatgpt-xu-hao.html)
• Chinchilla data-optimal scaling laws (https://lifearchitect.ai/chinchilla/)
• Jevons Paradox and LLMs (https://matt-rickard.com/jevons-paradox-and-llms)
• A Google AI model developed a skill it wasn't expected to have (https://qz.com/google-ai-skills-sundar-pichai-bard-hallucinations-1850342984)
• 4 Million Accounts Compromised by Fake ChatGPT App (https://www.artisana.ai/articles/4-million-accounts-compromised-by-fake-chatgpt-app)
• LLaVA: Large Language and Vision Assistant (https://llava-vl.github.io/)
• An open, billion-scale corpus of images interleaved with text. (https://github.com/allenai/mmc4)
• Connect to Google Colab VM locally from VSCode (https://github.com/amitness/colab-tunnel)
• OP Vault ChatGPT: Give ChatGPT long-term memory using the OP Stack (https://github.com/pashpashpash/vault-ai)

• Stability AI Launches the First of its StableLM Suite of Language Models (https://stability.ai/blog/stability-ai-launches-the-first-of-its-stablelm-suite-of-language-models)
• Exclusive: AI startup LangChain taps Sequoia to lead funding round at a valuation of at least $200 million
• “Langchain is Overrated”, and other strategies I use to deal with the GPT Hype (https://medium.com/@aplaceofmind/langchain-is-overrated-and-other-strategies-i-use-to-deal-with-the-gpt-hype-af06aa1a6be1)
• Autonomous Agents & Agent Simulations (https://blog.langchain.dev/agents-round/)
• GPT-3 Creative Fiction (https://gwern.net/gpt-3)
• Could ChatGPT become an architect? (https://constructionphysics.substack.com/p/could-chatgpt-become-an-architect)
• How to train your own Large Language Models (https://blog.replit.com/llm-training)
• Why universities should return to oral exams in the AI and ChatGPT era (https://theconversation.com/why-universities-should-return-to-oral-exams-in-the-ai-and-chatgpt-era-203429)
• Ideas to Professional Presentations in Seconds (https://www.magicslides.app/)
• How physics advanced Generative AI (https://www.assemblyai.com/blog/how-physics-advanced-generative-ai/)
• Google employees label AI chatbot Bard ‘worse than useless’ and ‘a pathological liar’ (https://www.theverge.com/2023/4/19/23689554/google-ai-chatbot-bard-employees-criticism-pathological-liar)
• The Anatomy of Autonomy: Why Agents are the next AI Killer App after ChatGPT (https://www.latent.space/p/agents)
• Teach GPTs to read API documentation using LLaMA, LoRA, and Langchain (https://github.com/danielgross/LlamaAcademy)
• web-llm: Bringing large-language models and chat to web browsers. (https://github.com/mlc-ai/web-llm)
• llama.go is like llama.cpp in pure Golang (https://github.com/gotzmann/llama.go)
• Leverage locally-hosted Large Language Models to write software + unit tests (https://github.com/catid/supercharger)

• Quora CEO Adam D’Angelo on Poe becoming the web browser for AI (https://www.semafor.com/article/04/19/2023/quora-ceo-adam-dangelo-on-poe-becoming-the-web-browser-for-ai)
• Snapchat making AI chatbot similar to ChatGPT available to every user (https://www.theguardian.com/technology/2023/ap19/snapchat-making-ai-chatbot-similar-to-chatgpt-available-to-every-user)
• Stack Overflow Will Charge AI Giants for Training Data (https://www.wired.com/story/stack-overflow-will-charge-ai-giants-for-training-data/)
• This new technology could blow away GPT-4 and everything like it (https://www.zdnet.com/article/this-new-technology-could-blow-away-gpt-4-and-everything-like-it/)
• How to run your own LLM (GPT) (https://blog.rfox.eu/en/Programming/How_to_run_your_own_LLM_GPT.html)
• The ChatGPT Artificial Intelligence Chatbot: How Well Does It Answer Accounting Assessment Questions? (https://publications.aaahq.org/iae/article/doi/10.2308/ISSUES-2023-013/10903/The-ChatGPT-Artificial-Intelligence-Chatbot-How)
• Introducing WebGPT (https://twitter.com/willdepue/status/1649147091573432321?s=20)
• Some Glimpse AGI in ChatGPT. Others Call It a Mirage (https://www.wired.com/story/chatgpt-agi-intelligence/)
• Google Looks to Turbocharge AI Efforts With Combined Brain, DeepMind Unit (https://www.wsj.com/articles/alphabets-google-combines-ai-research-units-brain-and-deepmind-620f4f7d)
• Experience CozoDB: The Hybrid Relational-Graph-Vector Database - The Hippocampus for LLMs (https://docs.cozodb.org/en/latest/releases/v0.6.html)
• The Embedding Archives: Millions of Wikipedia Article Embeddings in Many Languages (https://txt.cohere.com/embedding-archives-wikipedia/)
• Building A ChatGPT-enhanced Python REPL (https://isthisit.nz/posts/2023/building-a-chat-gpt-enhanced-python-repl/)
• Large Language Models: Scaling Laws and Emergent Properties (https://cthiriet.com/articles/scaling-laws)
• San Francisco GPT and CensusGPT — natural language interfaces to public data (https://github.com/caesarhq/textSQL)
• Text-prompted Generative Audio Model (https://github.com/suno-ai/bark)
• One step Baby AGI (https://github.com/Methexis-Inc/OneStepBabyAGI)
• Retrieve data from databases using natural language commands (https://github.com/logunify/dsensei)
• Runs a sharable ChatGPT docker container, allowing you to share a ChatGPT session with friends via a web browser. (https://github.com/BLuFeNiX/PlayGPT)
• Run Large Language Model "Agents" in Salesforce apex (https://github.com/callawaycloud/llm-apex-agents)

• Bard now helps you code (https://blog.google/technology/ai/code-with-bard/)
• ChatGPT could cost over $700,000 per day to operate. Microsoft is reportedly trying to make it cheaper (https://www.businessinsider.in/tech/news/chatgpt-could-cost-over-700000-per-day-to-operate-microsoft-is-reportedly-trying-to-make-it-cheaper-/articleshow/99637548.cms)
• SafeGPT: New tool to detect LLMs' hallucinations, biases and privacy issues (https://www.giskard.ai/safegpt)
• ChatGPT finds hundreds of vulnerabilities every day in NPM (https://twitter.com/feross/status/1648055850060374016)
• The inside story of ChatGPT's astonishing potential (https://www.ted.com/talks/greg_brockman_the_inside_story_of_chatgpt_s_astonishing_potential/no-comments)
• A snapshot of today's open-source LLMs space (https://twitter.com/theaievangelist/status/1645809824314298368)
• Yokosuka becomes Japan's first city to use ChatGPT for administrative tasks (https://www.japantimes.co.jp/news/2023/04/20/national/chatgpt-yokosuka-trial/)
• Google Colab no longer allows remote UI (https://twitter.com/artificialguybstatus/1649178893020979201)
• ChatGPT4 writes Stan code so I don’t have to. (https://statmodeling.stat.columbia.edu/2023/04/18/chatgpt4-writes-stan-code-so-i-dont-have-to/)
• We all contribute to AI – should we get paid for that? (https://techcrunch.com/2023/04/21/as-ai-eliminates-jobs-a-way-to-keep-people-afloat-financially-thats-not-ubi/)
• ChatGPT creates mostly insecure code, but won't tell you unless you ask (https://www.theregister.com/2023/04/21/chatgpt_insecure_code/)
• ProfileGPT: What does ChatGPT know about you ? (https://sahbichaieb.com/profilegpt/)
• Training open-source LLMs on ChatGPT output is a really bad idea (https://gist.github.com/mlaprise/bf4745655194162babfc2d158162e2e0)
• Human creativity in an AI saturated future (https://metastable.org/everything.html)
• Chart-GPT (https://twitter.com/whoiskatrin/status/1650245885127806976)
• WasmGPT ChatGPT-like chatbot in the browser(https://github.com/lxe/wasm-gpt/tree/wasm-demo)

• OpenAI has applied for “GPT” trademark with USPTO (https://news.ycombinator.com/item?id=35692476)
• openai Brand guidelines (https://openai.com/brand)
• Grimes Tells Fans To Deepfake Her Music, Will Split 50% Royalties With AI (https://www.forbes.com/sites/martineparis/2023/04/24/grimes-tells-fans-to-deepfake-drake-her-welcomes-collaboration-with-ai/)
• Large language models could improve their performance by outsourcing tasks to specialist AIs (https://spectrum.ieee.org/large-language-models)
• Bard vs GPT — Math, Logic and Problem Solving BATTLE (https://medium.com/nice-math-problems/bard-vs-gpt-math-logic-and-problem-solving-battle-%EF%B8%8F-ec8915fbd7e2)
• GCP: Supercharging security with generative AI (https://cloud.google.com/blog/products/identity-security/rsa-google-cloud-security-ai-workbench-generative-ai)
• ChatGPT now has a new Russian rival called GigaChat (https://www.businesstoday.in/technology/news/story/chatgpt-now-has-a-new-russian-rival-called-gigachat-378618-2023-04-24)
• RedPajama training progress at 440 billion tokens (https://www.together.xyz/blog/redpajama-training-progress)
• Transcribing Audio on AWS with Whisper: Cost Comparison AWS vs OpenAI (https://www.paulruales.com/blog/whiser-aws-guide)
• H2O LLM Studio - a framework and no-code GUI for fine-tuning LLMs (https://github.com/h2oai/h2o-llmstudio)
• LoopGPT Modular Auto-GPT Framework (https://github.com/farizrahman4u/loopgpt)
• Low-code LLM (https://github.com/microsoft/TaskMatrix/tree/main/LowCodeLLM)
• A GPT-empowered penetration testing tool (https://github.com/GreyDGL/PentestGPT)

• Hugging Face releases its own version of ChatGPT(https://techcrunch.com/2023/04/25/hugging-face-releases-its-own-version-of-chatgpt/)
• ChatGPT users can now turn off chat history, allowing you to choose which conversations can be used to train our models (https://openai.com/blog/new-ways-to-manage-your-data-in-chatgpt)
• The UK will spend £100 million to develop its own 'sovereign' AI (https://www.engadget.com/the-uk-is-creating-a-100-million-ai-taskforce-143507868.html)
• Google’s search business is feeling the heat (https://www.theverge.com/2023/4/25/23698087/alphabet-google-earnings-q1-2023)
• An AI Scraping Tool Is Overwhelming Websites With Traffic (https://www.vice.com/en/article/dy3vmx/an-ai-scraping-tool-is-overwhelming-websites-with-traffic)
• Amazon CodeWhisperer is out of beta. We tried it. Spoiler: it isn’t good (https://codeium.com/blog/amazon-codewhisperer-review)
• Web LLM (https://mlc.ai/web-llm/)
• Language Models Enable Simple Systems for Generating Structured Views of Heterogeneous Data Lakes (https://github.com/HazyResearch/evaporate)
• Layered, depth-first reading—start with summaries, tap to explore details, and gain clarity on complex topics (https://github.com/1rgs/clarity-reader)

• Mark Zuckerberg says Meta wants to ‘introduce AI agents to billions of people’ (https://www.theverge.com/2023/4/26/23699633/mark-zuckerberg-meta-generative-ai-chatbots-instagram-facebook-whatsapp)
• Yellow AI launches dynamic automation platform powered by generative AI for enterprises (https://venturebeat.com/ai/yellow-ai-launches-dynamic-automation-platform-powered-by-generative-ai-for-enterprises/)
• Google Researchers Unleash AI Performance Breakthrough for Mobile Devices (https://www.artisana.ai/articles/google-researchers-unleash-ai-performance-breakthrough-for-mobile-devices)
• Microsoft Releases SynapseMl v0.11 with support for ChatGPT, GPT-4, Causal Learning, and More (https://techcommunity.microsoft.com/t5/azure-synapse-analytics-blog/what-s-new-in-synapseml-v0-11/ba-p/3804919)
• Transformers from Scratch (https://e2eml.school/transformers.html)
• Store Embeddings in redis (https://release.com/blog/training-chatgpt-with-custom-libraries-using-extensions)
• FastGPT: A fast, full search, AI answering engine (https://labs.kagi.com/fastgpt)
• DeepDoctection: Document extraction and analysis using deep learning models (https://github.com/deepdoctection/deepdoctection)
• DeepFloyd IF: open-source text-to-image model (https://github.com/deep-floyd/IF)

• ChatGPT Grows in Popularity as Bing and Bard Flatline (https://www.artisana.ai/articles/chatgpt-grows-in-popularity-as-bing-and-bard-flatline)
• ChatGPT is finally ready to leave 2021. It can browse the internet now.
• (https://i.redd.it/its-official-chatgpt-is-finally-ready-to-leave-2021-it-can-v0-r6x7pu2m8bwa1.png?s=5fe6d2d520b1f13dfa95c2c90ef09720db5f1523)
• ChatGPT Prompt Engineering for Developers (https://www.deeplearning.ai/short-courses/chatgpt-prompt-engineering-for-developers/)
• Europe to ChatGPT: Disclose Your Sources (https://www.wsj.com/articles/europe-to-chatgpt-disclose-your-sources-863ef330)
• NVIDIA Announces NeMo Guardrails: An Open-Source Tool Designed to Improve the Performance and Safety of AI-Powered Chatbots like ChatGPT (https://www.marktechpost.com/2023/04/26/nvidia-announces-nemo-guardrails-an-open-source-tool-designed-to-improve-the-performance-and-safety-of-ai-powered-chatbots-like-chatgpt/)
• China Expert Gives Startling Reason Why CCP Is Requiring AI to ‘Reflect Socialist Values’ (https://www.newsbusters.org/blogs/free-speech/catherine-salgado/2023/04/27/exclusive-china-expert-gives-startling-reason-why)
• Unleashing Infinite-Length Input Capacity for Large-scale Language Models with Self-Controlled Memory System (https://arxiv.org/abs/2304.13343)
• LLM+P: Empowering Large Language Models with Optimal Planning Proficiency (https://arxiv.org/abs/2304.11477)
• Knowledge Retrieval Architecture for LLM’s (https://mattboegner.com/knowledge-retrieval-architecture-for-llms)

• Could you train a ChatGPT-beating model for $85,000 and run it in a browser? (https://simonwillison.net/2023/Ma17/beat-chatgpt-in-a-browse )
• HuggingGPT: Solving AI Tasks with ChatGPT and its Friends in HuggingFace (https://arxiv.org/abs/2303.17580)
• Make loading weights 10-100x faster (https://github.com/ggerganov/llama.cpp/pull/613)
• AI-enhanced development makes me more ambitious with my projects (https://simonwillison.net/2023/Ma27/ai-enhanced-development/)
• Italian privacy regulator bans ChatGPT (https://www.politico.eu/article/italian-privacy-regulator-bans-chatgpt/)
• Vocode–an open source library for building LLM applications you can talk to. (https://github.com/vocodedev/vocode-python)
• Join the Responsible AI Challenge (https://future.mozilla.org/builders-challenge/)
• Twitter recommendation source code now available to all on GitHub (https://twitter.com/elonmusk/status/1641876892302073875)
• Docker and Hugging Face Partner to Democratize AI (https://www.docker.com/blog/docker-and-hugging-face-partner-to-democratize-ai/)
• GPT-4 saves doggo’s life (https://the-decoder.com/gpt-4-saves-doggos-life/)
• 'He Would Still Be Here': Man Dies by Suicide After Talking with AI Chatbot, Widow Says (https://www.vice.com/en/article/pkadgm/man-dies-by-suicide-after-talking-with-ai-chatbot-widow-says)
• Interact ChatGPT via phone call (https://callsam.ai/)

• 30B model now needs only 5.8GB of RAM? How? (https://twitter.com/jeremyphoward/status/1642726595436883969)
• SudoLang: A Powerful Pseudocode Programming Language for LLMs (https://medium.com/javascript-scene/sudolang-a-powerful-pseudocode-programming-language-for-llms-d64d42aa719b)
• Adobe Firefly vs. MidJourney AI image generation (https://twitter.com/DrJimFan/status/1642921475849203712)
• State-of-the-art open-source chatbot, Vicuna-13B, just released model weights (https://twitter.com/lmsysorg/status/1642968294998306816)
• Introducing Agents in Haystack: Make LLMs resolve complex tasks
• Models guide (https://github.com/Crataco/ai-guide/blob/main/guide/models.md)
• Marvin: a batteries-included library for building AI-powered software. Marvin's job is to integrate AI directly into your codebase by making it look and feel like any other function (https://github.com/PrefectHQ/marvin)

• Stanford just released a MASSIVE 386-page report on the state of AI. (https://twitter.com/nonmayorpete/status/1643012248720523264)
• Large Language Model Observability (https://monitoring2.substack.com/p/large-language-model-observability)
• Goldman Sachs Predicts 300 Million Jobs Will Be Lost Or Degraded By Artificial Intelligence (https://www.forbes.com/sites/jackkelly/2023/03/31/goldman-sachs-predicts-300-million-jobs-will-be-lost-or-degraded-by-artificial-intelligence/?sh=a8d292b782b4)
• Believe Your Assignment is AI-Immune? Let's Put it to the Test (https://automated.beehiiv.com/p/believe-assignment-aiimmune-lets-put-test)
• Growth of AI Through a Cloud Lens (https://mitchellh.com/writing/ai-through-a-cloud-lens)
• Someone Added ChatGPT to a Furby (https://twitter.com/jessicard/status/1642671752319758336)
• you can greatly increase GPT-4's effective context size by asking it to compress your prompts using its own abbreviations. (https://twitter.com/NoLongerBennett/status/1642909851947810817)
• Microsoft Semantic Kernel Enables LLM Integration with Conventional Programs (https://www.infoq.com/news/2023/03/microsoft-semantic-kernel/)
• Open-source IDE powered by AI agents that do the work for you (https://github.com/e2b-dev/e2b)
• Converse with ChatGPT inside VS Code (BYO OpenAI API Key) (https://github.com/yagil/ChatIDE)
• Promptr is a CLI tool for operating on your codebase using GPT. (https://github.com/ferrislucas/promptr)

• CAMEL: Communicative Agents for “Mind” Exploration of Large Scale Language Model Society (https://github.com/lightaime/camel)
• Improving Large Language Models with External Knowledge and Automated Feedback (https://www.microsoft.com/en-us/research/group/deep-learning-group/articles/check-your-facts-and-try-again-improving-large-language-models-with-external-knowledge-and-automated-feedback/)
• AI is entering an era of corporate control (https://www.theverge.com/23667752/ai-progress-2023-report-stanford-corporate-control)
• Simulating an FTP server at ftp.disney.com with ChatGPT4 (https://www.filestash.app/2023/04/01/chat-gpt-acting-as-a-ftp-serve)
• OpenAI's approach to AI safety (https://openai.com/blog/our-approach-to-ai-safety)
• Segment Anything: model from facebook (https://ai.facebook.com/blog/segment-anything-foundation-model-image-segmentation/)
• Chinese Gaming Firm Appoints AI Robot As CEO, Sees 10 Percent Growth In Share Price (https://news.abplive.com/technology/chinese-gaming-firm-netdragon-websoft-appoints-ai-robot-as-ceo-sees-10-percent-growth-in-share-price-1589029)
• The Language Wars Are Over: ChatGPT Won (https://bourgoin.dev/posts/programming-languages/)
• JARVIS, a system to connect LLMs with ML community (https://github.com/microsoft/JARVIS)
• I’m an ER doctor: Here’s what I found when I asked ChatGPT to diagnose my patients (https://inflecthealth.medium.com/im-an-er-doctor-here-s-what-i-found-when-i-asked-chatgpt-to-diagnose-my-patients-7829c375a9da)

• The LLama Effect: How an Accidental Leak Sparked a Series of Impressive Open Source Alternatives to ChatGPT (https://thesequence.substack.com/p/the-llama-effect-how-an-accidental)
• Stanford benchmarks and compares numerous Large Language Models (https://crfm.stanford.edu/helm/latest/?group=core_scenarios)
• Fine-tuning with LoRA: create your own avatars & styles! (https://www.kix.in/2023/04/07/sd-lora-finetuning/)
• Thinking companion, companion for thinking (https://www.oneusefulthing.org/p/thinking-companion-companion-for)
• How ChatGPT is changing the job hiring process, from the HR department to coders (https://www.cnbc.com/2023/04/08/chatgpt-is-being-used-for-coding-and-to-write-job-descriptions.html)
• Harness the power of ChatGPT inside the GDB debugger! (https://github.com/pgosaChatGDB)
• DoctorGPT diagnoses your software's failures (https://github.com/ingyamilmolinadoctorgpt)
• Self-hosted AI coding assistant (https://github.com/TabbyML/tabby)

• GPT-5 arriving by end of 2023 (https://www.reddit.com/ChatGPT/comments/12gkt3z/gpt5_arriving_by_end_of_2023/)
• 5 Lessons Learned from Adding ChatGPT to a Mature Product (https://blog.trymito.io/5-lessons-learned-from-adding-chatgpt-to-a-mature-product/)
• ChatGPT with long-term memory (https://the-decoder.com/memorygpt-is-like-chatgpt-with-long-term-memory/)
• GitHub Copilot stealing other people's copyright comment blocks (https://news.ycombinator.com/item?id=35516887)
• Lost big client because of ChatGPT (https://www.reddit.com/freelanceWriters/comments/12ff5mw/it_happened_to_me_today/)
• LLMs won't eat the world-they will just homogenize it (https://www.vox.com/future-perfect/23674696/chatgpt-ai-creativity-originality-homogenization)
• Edit Anything by Segment-Anything (https://github.com/sail-sg/EditAnything)
• A curated list of awesome ChatGPT resources, including libraries, SDKs, APIs, and more (https://github.com/eon01/awesome-chatgpt)
• TurboPilot: self-hosted copilot clone which uses the library behind llama.cpp to run the 6 Billion Parameter Salesforce Codegen model in 4GiB of RAM. (https://github.com/ravenscroftj/turbopilot)
• Multi-task Language Understanding on MMLU (https://paperswithcode.com/sota/multi-task-language-understanding-on-mmlu)
• Telegram ChatGPT Concierge Bot (+ Voice!) (https://github.com/RafalWilinski/telegram-chatgpt-concierge-bot)

• Jailbreaking AI Chatbots Is Tech’s New Pastime
• OpenAI’s Bug Bounty Program (https://openai.com/blog/bug-bounty-program)
• GPT-4 Outperforms Elite Crowdworkers, Saving Researchers $500,000 and 20,000 hours (https://www.artisana.ai/articles/gpt-4-outperforms-elite-crowdworkers-saving-researchers-usd500-000-and-20)
• GPT-4 gets a B on my quantum computing final exam! (https://scottaaronson.blog/?p=7209)
• GPT-4 is highly susceptible to prompt injections and will leak its system prompt with very little effort applied (https://twitter.com/alexalbert__/status/1645909635692630018)
• Chinese tech giant Alibaba to release ChatGPT rival (https://www.dw.com/en/chinese-tech-giant-releases-chatgpt-rival/a-65275609)
• Generative AI-enabled compliance for software development (https://github.blog/2023-04-11-generative-ai-enabled-compliance-for-software-development/)
• Chunking Strategies for LLM Applications (https://www.pinecone.io/learn/chunking-strategies/)
• Maximizing the Potential of LLMs: A Guide to Prompt Engineering (https://www.ruxu.dev/articles/ai/maximizing-the-potential-of-llms/)
• Summary Of Adapter Based Performance Efficient Fine Tuning (PEFT) Techniques For Large Language Models (https://smashinggradient.com/2023/04/11/summary-of-adapter-based-performance-efficient-fine-tuning-peft-techniques-for-large-language-models/)
• GPT–LLM native macOS app with time travel, versioning, search (https://thellm.app/)
• Microsoft Kiota: CLI for generating an API client to call OpenAPI-described API (https://learn.microsoft.com/en-us/openapi/kiota/overview) (https://devblogs.microsoft.com/microsoft365dev/you-can-now-use-microsoft-kiota-within-visual-studio-code/)
• DeepSpeed Chat: Easy, Fast and Affordable RLHF Training of ChatGPT-like Models at All Scales (https://github.com/microsoft/DeepSpeed/tree/masteblogs/deepspeed-chat)
• Distributed² Machine Learning notebooks with Elixir and Livebook (https://news.livebook.dev/distributed2-machine-learning-notebooks-with-elixir-and-livebook---launch-week-1---day-2-1aIlaw)
• Large language model evaluation and workflow framework from Phase AI. (https://github.com/wgryc/phasellm)

• Dolly 2.0: Open Instruction-Tuned LLM (https://www.databricks.com/blog/2023/04/12/dolly-first-open-commercially-viable-instruction-tuned-llm)
• Building LLM applications for production (https://huyenchip.com/2023/04/11/llm-engineering.html)
• AI search engine for developers (https://www.phind.com/)
• Midjourney AI Guide (https://enchanting-trader-463.notion.site/Midjourney-AI-Guide-41eca43809dd4d8fa676e648436fc29c)
• Teaching Large Language Models to Self-Debug (https://arxiv.org/abs/2304.05128)
• Decreasing cold boot time in serverless GPUs(https://www.banana.dev/blog/turboboot)
• Use ChatGPT from WhatsApp (https://brightbot.chat/en/home)
• Yet another cache, but for ChatGPT (https://zilliz.com/blog/Yet-another-cache-but-for-ChatGPT)
• Excited by hype, CEO tells staff to use ChatGPT to build database of diseases (https://nwn.blogs.com/nwn/2023/04/chatgpt-hallucination-enterprise-use.html)
• I'm dating a chatbot trained on old conversations between me and my ex (https://www.reddit.com/OpenAI/comments/12696oq/im_dating_a_chatbot_trained_on_old_conversations/)
• Scaffolded LLMs as natural language computers (https://www.beren.io/2023-04-11-Scaffolded-LLMs-natural-language-computers/)
• Robot powered by GPT4 (https://twitter.com/DataChaz/status/1642558575502405637)
• ChatGPT – Who Owns the Content Created? (https://www.jdsupra.com/legalnews/chatgpt-who-owns-the-content-generated-2891692/) (https://news.ycombinator.com/item?id=35191206)
• Collection of LLM resources that can be used to build products (https://github.com/imaureawesome-decentralized-llm)
• OpenAI releases Consistency Model for one-step generation (https://github.com/openai/consistency_models)
• GPT4 gets a 0 on Steven Landsbrug's undergrad econ exam (https://twitter.com/ATabarrok/status/1646185259866939394)
• emergent abilities of large language models (https://www.jasonwei.net/blog/emergence)
• Praetor, a lightweight finetuning data and prompt management tool (https://github.com/gkamer8/praetor-data)
• Replacing my best friends with an LLM trained on 500,000 group chat messages (https://www.izzy.co/blogs/robo-boys.html)

• 1 Game map (More Coming soon)
• 5 Weapons
• A Few Achievements
• Many Custom Mobs (More to come with new maps)
• Permanent Upgrades with virtual currency got from games
• Rounds
• 1 Boss
• Different Armors

Top 5 Updates of last Week:

• Tiktok is testing a AI chatbot to improve user interactions with content.
• Pinterest introduced new platform change to improve their algorithm, the linked post is complex, I will be breaking it all down next week & share a post.
• Youtube is retiring Youtube Stories next month.
• Google added 3 new AI features to Search for all users to access.
• Twitter is asking researchers to pay $42k or delete data it gave them.
• Microsoft published new 41 page report on How to govern AI usage, promoting safe AI usages.

Google:

• Google expands its Digital marketing coaching program for SMBs.
• Google announced new AI marketing tools for SMBs and marketers.
• Google top story badge replaced with Trusted store badges.
• Google search bug caused a huge drop in video traffic.
• Google sitename now support subdomains on mobile.
• Google Bard gains images with more coming soon.
• Google Marketing Live 2023 Event recap for you.

Tiktok:

• TikTok sues Montana to prevent the app ban in the US State.
• Adweek’s new research in how tiktok creators are not getting the same reach as before.

Instagram

• Instagram Search Ads API is now available via Instagram marketing API.
• Instagram will soon allow you to add music to your profile.
• Instagram was down this week and caused people rushing to twitter.

Meta :

• Meta sells Giphy to Shutterstock.
• Meta slashes business teams in final round of layoffs.
• 1.2 Billion euro fine is now issued for Facebook as a result of EUPB binding decision.
• Meta introduces promotional ads for Ecommerce deals.
• Meta is about to limit how it uses Ad data in UK after UK antitrust forces them.
• Meta’s open-source speech AI recognises 4,000 global languages.

Twitter:

• Twitter bot situation is getting out of hands.
• Twitter is likely to opt out of EU code against disinformation.
• Twitter is working on a “video for you” section to promote video content.
• Twitter announced a new API pricing to help startups use Twitter at a reasonable flat rate.
• Twitter Adds new Playback speed control features for Video.

YouTube:

• Youtube shares their insights on Youtube algorithm for creators.
• Youtube launched Quiz posts on iOS and web.
• Youtube starts a weekly series on Intro to Live shopping.
• Youtube community posts feature is now available to all channels.

LinkedIn:

• LinkedIn Added 5 new Insights-backed features to Sales navigator.
• LinkedIn launched new security features: verified job posting & message alerts.

Snapchat:

• Snapchat reaches 200M Active users in India.
• Snapchat enhanced AR try-on tools as it continues to grow its AR capabilities.

Reddit:

• Reddit shared new insights about how new apps are discovered through daily App discussions.
• Reddit users are getting spam followers due to an ongoing spam campaign.

Advertising:

• Walmart working on providing Ad creative services to advertisers of all sizes.
• Lebron james joins Taco bell for “Taco Tuesday” trademark campaign.
• Disney & booking (com) making a huge spark with their new campaign.

Sales:

• Amazon’s Generative AI sales playbook got leaked.

Microsoft & AI

• Bing integrates itself with ChatGPT.
• Bing chat doubles chat character limit.

Pinterest:

• Many Platforms say they have Gen-Z using their app. But Pinterest shuffle is getting the real love.

Marketing:

• Discord gets custom streaming deck hardware and a big plug-in upgrade.
• Monday .com announces new integration with Microsoft teams.
• Neeva search engine, acquired by Snowflake after going private last week.
• MrBeast’s feastables brand is now available in every 7/11 in US.
• Peloton working on their new identity of more than a bike company.