OpenAI’s decision to not do closed source OpenAI Chief Scientist Ilya S. explains that OpenAI is not doing closed source because of safety reasons, but because of competitive reasons. He believes models can be improved to the point in the future where it becomes a safety concern. Competitive reasons The competitive reasons mentioned by Ilya S. refer to Microsoft, which is a major investor in OpenAI. Microsoft has been known to use closed source models in the past, which gives them a competitive advantage.

Introduction Group chat transcript on Generative AI Chaotic discussion on various topics Learning Resources Link to a post on what transformers are Cohere has two of the best ML content creators List of top AI-themed newsletters shared 42papers.com recommended Ben’s Bites newsletter recommended Hackathon Invitation to join a team in Warpspeed GenAI hackathon Deterministic Output Discussion on making GPT2 or BLOOM model outputs consistent Two ways to make it deterministic: setting temp=0 or setting seeds Pro tip to pursue this line of reasoning and gain first-hand experience Visual Aesthetic Scoring Discussion on AVA dataset and image dataset curation methods Link to a space on huggingface for aesthetic predictor Use of clip interrogator to score generated captions Hallucination and determinism are not related Visual aesthetic scoring is largely a final layer problem Discussion on the limitations of the model Dataset Creation Request for good library references to create datasets for LLMs Discussion on creating a dataset based on few hand-written examples Suggestion to check the terms of service for commercial use Discussion on using a visual QA model to identify errors in a picture Workflow suggestion to detect errors in the image and fix them using models finetuned on those datasets Suggestion to use Andrew Ng’s landing.

OpenAI and Google models: Google models have no moat OpenAI may have a moat due to early mover advantage and alignment with Microsoft/Azure OpenAI’s moat may come from government and finance domains OpenAI may be able to maintain its lead for 2 to 5 years or more Open source models are far behind OpenAI currently, but may catch up eventually Network effects may not be a moat in this case Microsoft+OpenAI can make inference costs low and onboard other businesses on plugins Hugging Face:

Introduction and greetings Members introduce themselves and exchange pleasantries Generative AI for digital painting Discussion on the potential for creating artist-focused digital painting tools using Generative AI Current methods involve hacking with automatic tools and using Photoshop to achieve desired results Large scale dataset creation for LLM model training Member asks if anyone has experience creating large scale datasets Links The description and link can be mismatched because of extraction errors.

Topics Discussed in Chaotic Generative AI Group Chat Transcript AI Models and Tools GPT4 creates a vector DB Replit open source LLM just dropped Finetune on instruction dataset curated from geeksforgeeks Prompt Injection in less than 10 minutes (video, slides, transcripts) JSONFormer: guaranteed JSON output with Huggingface LLMs DeepFloyd or Multi-ControlNet? Is there a model (other than gpt4) that can extract info from an image in JSON format? Autogpt app Table-transformer might help if result cards are in tabular form AI Applications Abstract artwork made using stable diffusion Extracting values without OCR Parsing information from result cards Langflow AI Industry and Innovation India’s position in the generative AI race Chris Lattner’s new ML-focused language Ecosystem graphs for AI startups Networking and Collaboration Connecting with others in the group for contribution and collaboration Looking for journalists or media folks in the group Miscellaneous Group invite link request AI and art techniques OpenAI service on Azure Multi-modal GPT4 access Sanford C.

Article Bias and Twitter Discussion about biased articles and preference for tweets Shared link to a tweet: https://twitter.com/zoink/status/1653052807950536706 Learning with ChatGPT Discussion on how to use ChatGPT for learning Mention of yudbot.com as a resource API Level Caps Mention of developing an API level cap for a llm vault manager Question on whether API level caps are generally kept Shared link to yudbot.com as a potential resource Paywalls and Capitalism Discussion on paywalls and access to news articles Mention of capitalism in a political group context Conversational Memory with GPT-3.

Access to GPT Plugins: OpenAI demos it publicly Participants want access to the tool Suggestion to make a Twitter viral demo to get access One participant has access but cannot find the tool in the UI Web Development: Request for freelance Flask/Python web stack developers Link to GitHub repository for Eva: https://github.com/georgia-tech-db/eva AutoGPT: AutoGPT is emerging as a catch-all for automation Jason Calacanis has posted a bounty for outbound sales emails on Replit One participant was able to deploy something locally to generate personalized DM, email, and coffee conversation points for a LinkedIn profile as a non-coder Group Management:

Music and AI Collaboration between GPT and Bark for music creation “Bark is lit too” 🔥 AI Tools and Libraries pandas-ai library: https://github.com/gventuri/pandas-ai Issues with ChatOpenAI giving “Could not parse LLM output:” errors Code snippets for language chain helpers and embeddings utils: https://github.com/jerryjliu/llama_index/blob/590639a14dd7346b7f5cc00a21dd24ce0d35ae30/gpt_index/langchain_helpers/text_splitter.py#L240 https://github.com/openai/openai-python/blob/c556584eff3b36c92278e6af62cfe02ebb68fb65/openai/embeddings_utils.py#L21 Scale.com for content language: https://scale.com/content-language AI Models and Projects Cohere’s multilingual model Project idea for creating shortcuts to automate tasks Resources for creating charts/graphs from data, open-source “Chart-GPT” resources especially python based Implementing file search using Langchain’s drive connector Performance gains attributed to Rust vs Golang for HNSW algorithm AI Discussions and Queries Issues with limited RAM on Repl on Replit Measuring recall for file search Tuning recall/speed for HNSW algorithm Learners’ discussion at 4pm Feedback on a new product for generating design from prompts using a design system Generative AI impact on graphic design and marketing Interviewing product designers and graphic designers in marketing agencies AI Resources and Links LLM.

VectorDB and LlamaIndex Discussion on whether LlamaIndex or Langchain support VectorDB with Sources with GPT4 or GPT3.5-Turbo Sharing of GitHub link for LlamaIndex demo Suggestion to use response.source_nodes to get sources Discussion on using evaluation module or regex to get desired sources Mention of OpenAI models not citing references Sharing of Langchain link for QA with sources example Discussion on limitations of Langchain and need to extend chain functions Mention of LlamaIndex not having such limitations AI Models and Applications Discussion on models for aesthetic score of images Mention of ResNet and classification model for user-scored images Discussion on generating professional-level DSLR photos using AI Mention of LAMINI AI library for fine-tuning LLMs to custom domains Announcement of upcoming “Learning Transformers/NLP/ML” discussion Discussion on D-ID and SadTalker for generating content Mention of Whisper models for Hindi transcription Suggestion of Deepgram and Monster API for transcription Discussion on building generative model for legible pdf/image documents Mention of AI-generated content for social media and advertisements Request for transcription API that can handle Hindi and regional languages Miscellaneous Discussion on automating extraction of WhatsApp group data Mention of RunwayML Gen 2 and Text2Video-Zero Request for transcription API that can handle Hindi and regional languages Announcement of weekly/monthly newsletter Request for tech expert in Generative AI for Zoom meet-up Request for connections with those working on advertisements Discussion on building AI avatars and personalized content Mention of guardrails for cloned voices Discussion on building QnA over CSVs using Python code or SQL Links The description and link can be mismatched because of extraction errors.

Generative AI Meetup Nvidia-HF event on Saturday morning Online event, URL: https://sites.google.com/huggingface.co/generative-ai-meetup Need an experienced person to guide and answer queries, volunteer needed Twitter link for volunteer: https://twitter.com/thesephist/status/1651677221797371904?t=UAtNw7WFH00_AS5oGpirUw&s=19 AI Applications Using diffusion for detection/recognition/segmentation tasks Tesla using diffusion as part of their lane detection algorithm Transformers used for lane detection, URL: https://youtu.be/aVjDX5XshYo Text2motion Promptlayer Data exploration and feature engineering resources needed Sequential prompting for controlled conversation, e.g. roleplay AlignmentAI and PMGPT for guided chat with roleplay AI assistant/co-pilot for product managers/product teams GPT plugins AutoGPT AutoGPT for basic questions AutoGPT’s limitations AutoGPT’s marketing Large scale human feedback and nudges for AutoGPT Namefinder.

Weaviate Weaviate allows pre-filtered vector search based on metadata and builds separate indices for metadata. Their managed SaaS charges based on the number and dimensions of vectors, irrespective of the extra metadata stored. Learning LLMs/Transformers/ML Someone suggested doing a zoom session to discuss learning LLMs/Transformers/ML. People discussed how they are dealing with rate limits and 5xx errors with OpenAI. Some suggested using a leaky bucket to smoothen the API calling rate.

OpenAI: Doubtful of claim that they don’t need data, they still need more data for training Multinationals may start using OpenAI models OpenAI is looking for data partnerships, but currently only able to cater to proposals from big tech companies No list of data partners available OpenAI’s platform API data policy is public information OpenAI is small and has exhausted internet and academic datasets for training their models Replit Codegen model is better than OpenAI Codex in many human eval tasks and is smaller in size OpenAI’s models are a utility like EC2 and have greatly improved the UX of consuming models Many NLP models now work with OpenAI’s transformer library Stanford NLP is highly regarded Qualcomm has made progress in ML on edge Palantir has launched ChatGPT for war Anduril may come up with something similar OpenAI’s tech use cases range from disaster relief to drone warfare Music and Audio Generation:

ChatGPT Chrome Extensions Discussion on cool ChatGPT related Chrome extensions RL without Human Feedback Shared interesting read on potential to use RL without human feedback OSS Projects Discussion on interesting OSS projects in the space ShareGPT Mention of ShareGPT AI Dating Expert Discussion on a bot trained on Bumble chats to reply like the user Prompt Injection Attacks Warning about prompt injection attacks Vector Search Discussion on using vector search on past swipes and bios OpenAI Branding Discussion on OpenAI branding and trademark on GPT Qdrant Discussion on using Qdrant for vector search Fine Tuning OpenAI Models Discussion on fine tuning OpenAI models on code Tips and do’s/don’ts for fine tuning Use of embeddings and GPT-3.

OpenAI and Dense Embeddings: Currently using OpenAI and dense embeddings Planning to move to hybrid soon Using similarity search on a flat embedding space to retrieve context GitHub and LLM: Came across a GitHub where someone used PEFT to fine-tune an LLM based on their iMessage chats to impersonate and create a bot that talks like them Supabase: Discussion about Supabase and Postgres performance queries Suggestions to check RLS policies and use explain.

Anthropics pitch deck Does anyone have access to Anthropics pitch deck? RLHF for instruct models OpenAI’s John Schulman gave an interesting talk at Berkeley last week on why RLHF was needed to get the instruct models to behave nicely. Deepfakes and audio synthesis First instance of an Indian politician referring to deepfakes / audio synthesis? AI courses and resources course.fast.ai for being able to make sense of all of this — even as it changes in 2-3 months and we add VQA (Vision) to mainstream OpenAI APIs Lot of new work should come from STT and TTS side, including performance improvements like Whisper-JAX in the coming 4-6 month and more important, voice cloning, avatars and the like.

Cloud GPUs: Discussion on using GPUs for training models Google Colab and Kaggle Notebooks suggested as options Comparison table for cloud GPUs shared Cheapest option for model fine tuning mentioned as rtx5000 on runpod.io Pricing of cloud GPUs discussed, including economies of scale and interest rates on capital investment Vast.ai mentioned as a cloud GPU rental option Model training: Tips shared for reducing container image size when using Sentence Transformers for word embeddings Question asked about enabling multiple checkpoints in automatic1111 while fine tuning a model using dreambooth Privacy concerns:

Web and AI Tools: Langchain and WebGPT discussed Kubiya.ai resource shared and discussed Helm charts and K8s clusters can be managed via chat interface using natural language Discussion on the evolution of computer engineering from command lines to GUIs and back to command lines AI Techniques: Advanced stable diffusion techniques for inpainting and controlnet discussed Tutorial on creating consistent AI characters across images with SD shared Request for AI art/text to image group Proposal for a separate DeepMedia: Generative Art (Text to Images, Video, Music) group Illustrated Transformers and other resources shared Pinecone:

Meet-up: Generative AI meet-up in BLR on Saturday evening Link: https://hasgeek.com/generativeAI/april-meetup/ LLMs: Replit’s blog post on training their own LLMs Link: https://blog.replit.com/llm-training OpenAI Whisper had a team of 6 for their project Attention is All You Need paper had 8 authors Midjourney outsourced frontend to Discord Microsoft’s LayoutLMv3 does OCR Link: https://huggingface.co/microsoft/layoutlmv3-base MM-REACT uses reasoning capabilities of LLMs to extract information from visually rich documents Link: https://github.com/microsoft/MM-REACT Tools: Best practices for recording LLM experiments W&B is a familiar tool CohereAI tweet about model specificity and accessibility Link: https://twitter.

UI Design: Used UI elements from Koo app and dreamstudio Gradio space and playgroundai.com suggested as starting points Jadoosnap.com suggested for product design Data Analysis: Shapiro Wilk test explained as measure of normal distribution Normality not necessary for regression, other methods suggested Support vector regression and tree-based methods suggested for non-normal datasets Extreme Value Theory tools suggested for non-normal datasets Hill Estimator suggested for non-normal datasets Image Editing: Inpainting models suggested for image editing Qdrant and Chroma suggested for custom databases Vall-e and Descript suggested for voice cloning DeepSpeed evaluated and recommended Ada from OpenAI and sentence transformers from HuggingFace compared Language Models:

OpenAI Plugin Store: Adept.ai got copied as an in the OpenAI Plugin Store, called Multi-on Experience needed: People with experience in CLIP, BLIP-2, and VQA are needed for a project Someone with solid experience in multimodal vector similarity search is also needed Meeting invite: A meeting invite was shared in the WhatsApp group for the project Projects: Multimodal vector similarity search Personal search Generative AI for solving problems and building multi-modal systems Domain-specific feedback loop and fine-tuning Context-based search with cohere embeddings and OpenAI GPT3.

Langchain and Mendable.ai integration for building with Langchain; Kapa Langchain bot for asking doubts on Discord; Paradox.ai for recruiters; Skillate and Leoforce for automated resume screening; issues with bias and possible unethical use of AI solutions in recruitment; use of GPT during interviews; AI-generated music; discussion on GPT-4 and its performance on JEE questions; fine-tuning vs. prompt engineering davinci; use of LLMs for math problem solving and chain-of-reason prompting; LOL dataset for low light pictures; Huggingface’s PEFT wrapper for multiple fine-tuning methods; founder’s R&D work for Bewgle; LLM components for memory weighing in Langchain; publishing research papers on generative AI; controllability issues with LLMs and use of guardrails for structured outputs.

Sci-Fi and AI Discussion on sci-fi authors such as Asimov, Clarke, and Heinlein Comparison of Multivac from Asimov’s stories to today’s LLMs/Auto-GPT Sharing of a sci-fi story by Greg Egan Multimodality and its potential to improve LLMs’ understanding of reality Sharing of a paper on color clustering Appreciation for Asimov’s work and genius Recommendation of Cixin Liu as a favorite author Community Guidelines Drafting of community guidelines for the group Sharing of a website for sci-fi enthusiasts (Orion’s Arm) GPT-Related Topics Request for a service that allows using chatGPT interface with API key Sharing of a website for chatbot UI Discussion on context window handling of such services Recommendation to use GPT-4 with large context windows for cost efficiency Request for guidance on fine-tuning with custom data and sharing of a GitHub example Sharing of articles by Databricks on tuning Dolly LLM Discussion on deploying a Docker image with Sentence Transformers and OpenAI combined Recommendation to deploy the model separately behind a service endpoint to avoid distributing the model in application containers Discussion on the use of embeddings and OpenAI for cost reduction Sharing of a Twitter thread on fine-tuning and prompting Recommendation to use external vector databases like Pinecone or Weaviate instead of fine-tuning for cost efficiency Discussion on chunk size for converting text dump and documents into embeddings Sharing of Langchain’s conversation memory types for optimization Question on the possibility of specialized hardware speeding up embedding search Request for a summary of the GPT-related topics due to the increasing number of messages in the group Sharing of a post by Vespa founder on introducing embeddings and vector search Sharing of an IPython ChatGPT extension Discussion on Auto GPT’s intermediate summary phase for webpage parsing Experience with LLMs and plans to add new tasks in classifier and separate their workflow standalone Sharing of Langchain’s memory chains and conversation memory for summarization and retrieval Links The description and link can be mismatched because of extraction errors.

React Development: Someone is looking for React developers for a paid weekend project They ask for leads to be DM’d to them Controlnet Inpainting: Someone asks if anyone has gotten the latest controlnet v1-1 nightly release models working with inpainting They mention that they have gotten the models working with multi controlnet, but not for inpainting yet Azure Cognitive Search: Someone brings up MSFT’s recommendation to use Azure Cognitive Search with Azure OpenAI for enterprise document search in Azure They mention that it may be costly and ask for thoughts from the community on production level deployments Another person mentions the option of uploading document embeddings into a Vector DB and doing semantic search They share that they are currently using this method with Pinecone and have 10681 vectors in their DB GPT-5:

LangChain and Tenacity API LangChain uses Tenacity API +1 on using Tenacity API LangChain has an API for the idea of kids as superheroes Andrej Karpathy’s Neural Nets Video Finished watching Andrej Karpathy’s video on Neural Nets Highly recommended for anyone starting or in the field Link to the video playlist: https://youtube.com/playlist?list=PLAqhIrjkxbuWI23v9cThsA9GvCAUhRvKZ Image Inpainting and Diffusers Tried Img2img models for inpainting on an online image Asked for suggestions on using diffusers for inpainting Link to Stable Diffusion XL beta: https://stability.

Dolly 2.0 Databricks has released Dolly 2.0, a commercially viable LLM. The dataset was crowdsourced from Databricks employees. The training code, dataset, and model weights have been open-sourced and are suitable for commercial use. Link: https://www.databricks.com/blog/2023/04/12/dolly-first-open-commercially-viable-instruction-tuned-llm OpenAI’s Consistency Models OpenAI has proposed consistency models, a new family of generative models that achieve high sample quality without adversarial training. Link: https://www.marktechpost.com/2023/03/10/open-ai-proposes-consistency-models-a-new-family-of-generative-models-that-achieve-high-sample-quality-without-adversarial-training/ Vector Databases Participants discussed various vector databases, including Pinecone, Weaviate, Qdrant, Chroma, Redis Vector Cache, Vespa, pgvector on Supabase, and Milvus.

AI Embedding Models: Comparison between OpenAI’s “text-embedding-ada-002” and Huggingface models or Sentence Transformers Some prefer Huggingface or Sentence Transformers for their use case Recommended model: “sentence-transformers/all-mpnet-base-v2” or “multi-qa-mpnet-base-dot-v1” for semantic search Cross-encoder recommended for better scoring mechanisms AI in News Media: Traditional News Media is frightened with AI Buzz is hot and controversial Interviews conducted on AI opinions AI Chatroom Plugins: ChatGPT Plugins SF hackathon winners announced Winners have higher usability than expected Indians are no longer playing catch up AI Conversational Memory:

Generative AI Discussion about a paper on generative agents and a demo at https://reverie.herokuapp.com/arXiv_Demo/ Mention of Andrej Karpathy’s tweet Sharing of a link to a website that uses flat GPT output without crawling their pages Mention of a bot that doesn’t have memory and a classic NER ambiguity mistake Discussion about ChatGPT plugins and OpenAI account access Speculation on how OpenAI determines a user’s country and suggestions for contacting OpenAI Mention of Ojasvi leading AI at mydukaan.

Hackathon Ideas Discussion about a thread of hackathon ideas and similarities to their own hackathon Political action project was the first thing they sold to a client back in 2022 Creators of the thread haven’t been mentioned, but can tweet to Joseph or @swyx to ask Mention of knowing @swyx from the react world because of a talk Variety of chatbots in their hackathon GPT 3.5 Discussion about problems with limiting output length Question about changing length units to token instead of characters Recommended way to do this is to use Guardrails and turn on re-ask Mention of Guardrails creator doing a demo at their hackathon venue Link to Guardrails GitHub page Mention of Indians no longer playing catch up in the tech world Kor and Guardrails Discussion about Kor being a general purpose parser for any text/schema Confusion about why the schema format in Guardrails is not the same as the output format Balancing language/lib design POV Resizing prompts work well when input and output language is English BPE tokeniser is notoriously unstable for CJK and Indian languages Prompt engineering is like casting a spell Suggestion to tweet and tag OpenAI folks for more information on GPT4 GPT4 and GPT3 Discussion about customers raising support tickets for being fooled about GPT4 access Suggestion to improve with a system prompt Chat models are instruction finetuned, sanitised and scaled forks of large base models Suggestion to have GPT3.

SAAS Company Sale Discussion about a tweet by Dan Martell regarding selling a SAAS company. Generative AI Tools Discussion about a tool for generating desi lofi girls. Mention of a tool built by [PHONE REMOVED] and friends for YouTube videos. Mention of sitegpt.ai, which does something similar. Mention of Microsoft Edge’s copilot option for generating text. Mention of jsonlines format for solving text generation challenges. Vector DBs Discussion about starting a vector DB company.

Integration of chatGPT/LLMs into smart home devices Query about integration of chatGPT/LLMs into Alexa/Nest type of smart home devices Mention of tools that can do it, but openai key is a barrier to adoption Paddlespeech Query about anyone dabbling with paddlespeech ChatGPT talk by Anil Ananthaswamy Video of the talk by Anil Ananthaswamy on ChatGPT held at BIC is published Link: https://youtu.be/WF28ZwhUCc4 Automatic1111 and DrawThings App Suggestion to use Automatic1111 on colab Suggestion to download DrawThings App on MacBook GPU memory issue, try smaller models Generative AI on local machine Discussion about running generative AI workflows locally on a Mac Recommendation to either go all the way or get the Macbook air Suggestion to get as big of a machine as your wallet permits Mention of text-to-video taking off in the future Optimization of models for running on Apple RAM Query about how optimized the models are for running on Apple RAM as opposed to Nvidia GPUs Serverless Vector DB and pricing Discussion about serverless Vector DB and pricing around it Link: https://twitter.

ViewsAct Project Request for demo video Creator opted out due to use of real-world company designs and data Hackathon Demos Discussion on winners in SF/NYC/Berlin circuits Request for more information on tracking Dalle Short Film Link shared: https://www.instagram.com/reel/CqstiaNuSjX/?igshid=MDJmNzVkMjY= Another link shared for breakdown process Networking Request to meet people working in gen AI space in Bay Area Introduction of Applied Scientist at Amazon building computer vision tech for video action recognition AI/Art/Gaming Event in London Link shared: https://www.

General AI Discussion: Speculation on how the system prompt is used in AI chat Discussion on building AI applications for schools Query on the best model to use for Sparse Vectors and hybrid search Experimentation with AI chat Query on the tool used for text2vid Query on Danny Postma’s cloth transfer technique Sharing of recent research on cloth transfer Sharing of a GitHub link for HR-VITON Query on MidJourney subscription payment Discussion on using Google Sheets as a database for GPT4 app Self-promotion of Gooey.

OpenAI Embeddings and Graph Embeddings Discussion on utilizing OpenAI embeddings model to create graph embeddings Naive approach suggested: use “text-embedding-ada-002” to get node embeddings and compute edge embeddings using adjacency matrix Pinterest recommends starting with any base embedding and then updating them Article on HN shared for reference GPT Model for Financial Data and Analysis Query on availability of GPT model finetuned with financial data and analysis Mention of Bloomberg GPT, but not yet usable or publicly available Doubtful if it will ever be open sourced as it is trained on proprietary data Generative AI Mixer Event Invitation to a generative AI mixer event at Lightspeed office in Koramangala Interested participants asked to fill up a Google form Deepfake Detection Query on research happening in deepfake detection and any good app/software for it Mention of Spoofsense.

Generative AI Tools and Platforms OpenPlayground and Playground.ai are discussed as generative AI experimentation platforms OpenPlayground is noted for its ability to experiment with different diffusion models, parameters, inpainting, and base image generation for LLMs Playground.ai is noted for its popular prompts and upvoting system API access for Playground.ai is expected to be released soon The group discusses the strategy of having a Discord bot but not an API, with some suggesting that it may be due to the cost and difficulty of setting up an inference infrastructure Vast.

Emergence of Creativity and Intelligence Creativity and human intelligence may just be an emergent property of a small number of algorithms operating with a lot of compute power. Many respected neocortex researchers believe there is effectively one algorithm for all intelligence. Human brains don’t look all that different from chimp brains, and yet somehow produce wildly different capabilities. Our own intelligence may just be the emergent combination of a bunch of cheap tricks.

Topic 1: Alternatives to D-ID and ElevenLabs User is looking for good/free/local alternatives to D-ID and ElevenLabs Wants to upload custom image and audio to get a lip-synced avatar video Gooey.ai/lipsync is suggested as a possible alternative Topic 2: Differences between Dreambooth, LoRA, and Textual Inversion User is looking for information on the differences, pros, cons, and training time of Dreambooth, LoRA, and Textual Inversion User has experience training a custom DB model but hasn’t used the other two yet Another user suggests that LoRA works well compared to the other two, especially after being trained on a small amount of data Dreambooth provides decent results after multiple attempts of fine tuning Textual Inversion didn’t return any good results for the user who tried it A YouTube video (https://youtu.

DeepHack Demo Day Discussion about who is attending the DeepHack demo day Live stream link shared: https://has.gy/ULYt Live stream temporarily stopped for a 5-minute break Submissions Question about whether all submissions have been completed Social Links None mentioned. Links The description and link can be mismatched because of extraction errors. https://has.gy/ULYt - Live stream set up for DeepHack demo day.

Quick Hacks and Extensions: A quick hack for DeepHack GPT to handle multiple context questions A small flex of exporting group chat from WhatsApp and uploading the text file to gooey for search without writing any code Whisper API can be used to get the transcript without GPUs Gladia.io can extract 1-hour audio file transcript in 10-20 seconds using openai Whisper backend Assembly.ai has a playground to upload files for transcription with good accuracy HuggingGPT: Solving AI Tasks with ChatGPT and its Friends in HuggingFace LLMs and Table Structures:

OpenAI and Tech Hindrance: Discussion on OpenAI’s decision to stop developing new AI models due to potential job loss Some members believe hindering tech is not fruitful OpenAI is a for-profit company backed by Microsoft Elon Musk’s involvement in OpenAI is questioned Yann Lecun’s stance on the issue is mentioned AI Models and Tools: Discussion on various AI models and tools such as GPT-4, ChatGPT, Bard, and Pix2struct Mention of MM-React and Llama Multimodal Interest in Karpathy’s work and Tesla’s use of synthetic data and big datasets Personal Experiences:

Google and Replit partnership to take on Microsoft’s GitHub India Replit sponsoring hackathon GCP features on Replit Discussion on popularity of writing GPT from scratch Request for safety protocols for AI development Comparison to past technologies and societal effects Proposal for philosophy room and use-case specific groups Azure OpenAI credits discussion Hacks to reduce GPT hallucination BIC event on ChatGPT and its ilk OpenAI grant eligibility and phishing concerns Future of Life open letter on pausing giant AI experiments Score system to penalize GPT hallucination Verification of GPT sources and self-consistency check Throttling using prompts discussion Proposal for custom agent implementation Self-consistency empirically known to work Links The description and link can be mismatched because of extraction errors.

Introduction: Dr. Parth Sharma introduces himself as a doctor with an interest in AI and deep learning. AI and Deep Learning: Discussion about Andrej Karpathy’s videos on YouTube. ChatGPT is used to generate a list of Shopify websites in the USA that sell candles. ChatGPT is also useful for competitor analysis. The group discusses the use of checkpoints and Exif data in AI image generation. Dreamboothing is mentioned as a technique for generating AI images.

Deep Hack Demo Day Invite Invitation to a demo day Sales pitch to share with friends in Bengaluru One member wishes they could attend but will be flying out that day Jobot, a ChatGPT-powered bot is live at Jovian Announcement of a new ChatGPT-powered bot One member asks if the demos will be streamed Introductions One member introduces themselves as an anthropologist and senior fellow with the Max Planck Institute for Social Anthropology Another member introduces themselves as a new media artist and educator, with a focus on using AI, robotics, and code to build experiences A third member introduces themselves as a search engineer at Wolfram|Alpha AI Projects and Demos Discussion of a video project that uses OCR, content search, style transfer, and projection mapping Sharing of a GitHub repository for a blind visual aid using GPT Discussion of a demo for a work-in-progress project that integrates ControlNet to alter per specific requirements for a particular color-blind person Sharing of a news article about a project that uses goggles for the blind Discussion of using serverless GPUs and accelerate to offload to CPU Discussion of using GPT to write and create episodes for the final season of Game of Thrones Discussion of GPT’s ability to handle degeneracies and reason about problems in its distribution but not in its data Discussion of GPT’s performance on new Leetcode problems past its training date Sharing of a blog post about cheating in coding using LLM-powered assistants Discussion of the potential for LLM-powered coding assistants to automate much of the code we write and the skills that will become more important as a result Miscellaneous Discussion of a text-to-video tool for camera movements over still images Sharing of a ControlNet web app and API Discussion of the eligibility of remote submissions for a prize Links The description and link can be mismatched because of extraction errors.

Video on OpenAI’s Development Trajectory Link: https://www.youtube.com/watch?v=L_Guz73e6fw Discussion about OpenAI’s development trajectory No technical details, but philosophical nuggets shared by Sam Altman Talks about the potential of AI and its impact on society Mentions the importance of safety and ethical considerations in AI development The AI Dilemma Video Link: https://www.youtube.com/@aajtakai/videos Discussion about the recent blitzscaling of OpenAI and Microsoft in AI advancements Video titled “The AI Dilemma” talks about the safety concerns and risks associated with exponential growth of LLMs (large language models) Emphasis on the need for ethical considerations and safety measures in AI development Links The description and link can be mismatched because of extraction errors.

Topics Discussed in Chaotic Generative AI Group Chat Transcript: AI Art Generation Workflow Discussion on using Stable Diffusion for AI art generation. Workflow involved using text2img with controlnet depth map of the image. Depth map was obtained from Clipdrop’s API. Automatic1111 was used for the entire process. Iteration loops were run in every step until satisfied with the output before moving to the next step. Retakes are common in AI art generation.

AI Model and Prompt Discussion about a prompt and AI model Need for composition iterations to achieve desired output Mention of a Rembrandt painting as a style reference for the second picture Mention of a custom AI model created by merging different models OpenAI Plugins Mention of OpenAI ChatGPT plugins Comparison to Alexa Skills Link to OpenAI Plugins documentation Off-Topic Discussions Request for help with Firestore Invitation to showcase projects Sharing of a link to an article about the cost of ChatGPT for Indian languages Discussion about using OpenCV videocapture() with an external iOS camera Support for direct vernacular interaction with GPT, but with poor performance.

Introduction One member introduces themselves as Dukaan’s head of AI. JadooSnap JadooSnap is a website that some members have seen and praised. It is not yet available for public access, but some companies are already using it. A member shares a link to JadooSnap.com. Access to JadooSnap A member asks if there is any way for Dukaan to request access to JadooSnap. Another member shares a Twitter link that may be helpful.

Introductions and Backgrounds: Vishal Tripathi, Venture Investor, worked on a fund of funds strategy for Google Ventures - Plexo Capital. Currently with Legacy Venture, investors in many of the funds people have mentioned above (Sequoia, A16z, Matrix, Accel, True, etc). Someone mentioned Yohei (from untapped VC) above, he’s a close friend and has collaborated with him on some GPT tools. Proof of Concept Script: A member has a proof of concept script that they would like to take to customers.

Generative AI Group Chat Transcript Product Inpainting: Discussion on using separate LoRA for inpainting main product Tutorial using ComfyUI for generating similar images with different ethnicities and anime styles of same character shared Clarification on main product being matchstick box with person holding it Flair and Booth.ai tools discussed for retaining text and product fidelity Better methods like ControlNet, Adapters, Composer suggested for similar use cases as Flair Finetuning models not needed for controlling style, image composition, and color scheme Copy-pasting pixels on final image where product exists suggested for preserving product fidelity Background subtraction and improvement suggested for e-commerce images Adept.

AI Models and Tools: Hugging Face’s text-to-video synthesis model: https://huggingface.co/spaces/damo-vilab/modelscope-text-to-video-synthesis Reid Hoffmann’s essay on GPT4 and AI summary request Midjourney for generic imagery and potential API release Stable Diffusion vs. Photoshop and Canva vs. Midjourney comparison InvokeAI’s open source UI for Stable Diffusion and potential API release Automatic1111 as a potential future option for CreativeAI RunwayML Gen1 and its impressive features Comfy UI, a node-based/DAG interface for image generation OpenAI’s paper on the potential impact of Large Language Models on the Labor Market: https://twitter.

Introduction Confirmed FOSS Demos for the hackathon, 2 of the top 3 projects from Github Trending in AI Thanks for the intro Ravi Theja, GPT-Index contributor and inMobi DS [PHONE REMOVED] to Jerry Liu Interesting Links Sharing my favourite blog from last few weeks: https://simonwillison.net/2023/Mar/17/beat-chatgpt-in-a-browser/ https://mobile.twitter.com/nickfloats/status/1631346749297106958 came across this today AI Models WASM for LLMs is super interesting, anyone tried? They also trained an LLM to mimic Homer Simpson and it’s apparently better than gpt-4 It’d be hilarious if the model forgot how to do chain of thought reasoning because of that 😂😂 Research Labs Iit madras or Bombay has research labs you can cold mail them or reach out to other iits they might help or redirect to appropriate org.

New BingGPT Launch: Potential to make advertising briefs easier Link: https://www.bing.com/new AI and Software Engineering: Discussion on how GPT4 will change software engineering Link to article: N/A March Meetup: Question about whether the March meetup for gen ai is happening No meetup in March, suggestion to come to the hackathon instead RSVP/sign up link requested Link: N/A Image Generation Models: Discussion on using image generation models Mention of Midjourney, cvitai, and openjourney models Recommendation to try gooey.

Language Models and Datasets Discussion on why language models have a cut off in 2021, including GPT-4 Training involves curating a dataset which is time-consuming Cleaning and wrangling data takes a lot of effort Some open source datasets are available GPT-4 paper might have the dataset details, but was disappointing Model has access to data from 2022, but official position is that the data has a cutoff date in 2021 AI Tools and Applications Discussion on the performance of AI tools for large scale searches Pinecone is recommended for specific vector databases ChatGPT-3 Whatsapp bot can summarise videos, transcribe voice notes, answer text questions, and generate images using /image Replit.

Paper Reading Groups Discussion about reading and analyzing research papers related to popular models Suggestions to play a game of hunting for clues and trying to understand the research team’s thought process Recommendation to start with interdisciplinary papers like Visual ChatGPT and diff it against other models Idea to host a reading session and invite people to present different papers and discuss insights Mention of CV paper reading group in Bangalore run by Sumod Mohan and a CS paper reading group run by Swanand via Hasgeek Interest expressed in listening to reading sessions Events and Resources Request for information about paper reading groups happening in Bangalore Suggestion to spontaneously organize such events in Bangalore Sharing of a Twitter link to an event by LangChainAI Discussion about storing embedding vectors in a PG database and the need to explore other options Links The description and link can be mismatched because of extraction errors.

OpenAI Demo and Productivity Tools: OpenAI demo of wireframe to working code was impressive Discussion on how it could be used for productivity tools like email, document writing, and slide creation Some productivity and note apps may become irrelevant after this technology Deep Learning/NLP: Discussion on various resources for learning about deep learning and NLP, including courses and papers Recommendation to watch Karpathy’s nonogpt video Discussion on studying existing models and their building blocks Xavi Amatriain’s catalog may be overwhelming for beginners Non-technical questions are welcome Transformers and Fine-Tuning:

Introduction Group chat discussing Generative AI Non-techie member seeking help to keep up with the discussion Participants [PHONE REMOVED] [PHONE REMOVED] [PHONE REMOVED] are PMs [PHONE REMOVED] [PHONE REMOVED] are founder and VC respectively [PHONE REMOVED] is Mumbai based creative agency person Manjot [PHONE REMOVED] is with Lightspeed VC currently Hackathon Prize money split is 1L minimum per track, split between all members of the winning team equally Evaluation criteria is left to the jury’s discretion Some members prefer leaving the criteria to the judges while others prefer having a clear set of criteria Tracks are announced for the hackathon Alpaca AI https://alpaca-ai-custom4.

OpenAI LLMs and Foundry Discussion about whether there are any LLMs by OpenAI that can be deployed on-premises today Mention of OpenAI Foundry, which costs $1 million per year and will have on-premises capabilities No specific LLMs mentioned that can be deployed on-premises currently FLan-T5 by Google Mention of FLan-T5, which is a FOSS (Free and Open Source Software) by Google Thank you Thanking Nirant for providing quick help Links The description and link can be mismatched because of extraction errors.

Hackathon Registration Participants only need to fill the Google form for the hackathon This chat is for everyone interested in seeing what was built during the hackathon Register for a hackathon invite to the BLR venue: https://forms.gle/UizUwyKi4bajt6WB7 GPT Hype Discussion about the hype surrounding GPT Question about whether it is different from the Google form for the hackathon Company Participation Discussion about the lack of company participation in the hackathon Someone from the company is encouraged to sign up to participate Runwayml Question about whether anyone has access to Runwayml No further discussion or context provided Links The description and link can be mismatched because of extraction errors.

Introduction Group chat about Generative AI Chaotic conversation Indian Generative AI Landscape Shared link to article about Indian Generative AI Landscape Discussion about the state of AI in India Mention of Indian startups in the AI space Mention of Indian government’s AI initiatives GPT-4 Shared link to Reddit post about GPT-4 Discussion about GPT-4 and its potential capabilities Speculation about the release date of GPT-4 Miscellaneous Thanked member for adding to the group chat No other specific topics discussed Links The description and link can be mismatched because of extraction errors.

Introduction A member of the group shared an opportunity with the group The opportunity is with a large content and consumer brands company in India The company reaches about 40% of all social media users in India The member asked if anyone in the group or anyone they know would be interested in exploring the opportunity No further discussion There was no further discussion on the opportunity shared by the member.

Generative AI Hackathon Participants discussing a Generative AI Hackathon No constraint on the problem statement April Fool Apps are welcome Participants asking if everyone has a team Normalizing Flows Participants discussing Normalizing Flows One participant has been interested in it for a while Another participant shows interest Discussion on industrial use cases or academic interest Resource Participants share a resource: https://research.runwayml.com/gen1 Links The description and link can be mismatched because of extraction errors.

Research and Breakthroughs Oxford University researchers introduce a diffusion model called RealFusion that can generate 360-degree reconstructions of objects from an image. [https://www.marktechpost.com/2023/02/28/oxford-university-researchers-introduce-a-diffusion-model-called-realfusion-that-can-generate-360-degree-reconstructions-of-objects-from-an-image/?amp] Breakthrough for nerf NeRF and 3D Art Devanshu Tak is a 3D artist who has used NeRF earlier in his work Interesting read: https://unsupervisedlearning.substack.com/p/using-large-language-models-effectively. Esp the #4 point around embeddings Embedding and Prompt Templating Embedding was the biggest reason for using Pinecone and Langchain in the Hasura bot Pinecone is used as a vector store Langchain gives a good abstraction Hasura bot uses openAI embedding Chroma bundles store and the embedding call into one from what can be told.

Introduction Anirudth introduced himself as an Applied Scientist at Amazon He expressed his excitement about generative AI and looking for new opportunities Generative AI Discussion about the potential of generative AI in various fields such as art, music, and literature Mention of OpenAI’s GPT-3 and its capabilities Sharing of articles and research papers related to generative AI Discussion about the ethical implications of generative AI and its impact on society Projects and Tools Sharing of personal projects related to generative AI Discussion about various tools and frameworks used for generative AI such as TensorFlow, PyTorch, and Keras Sharing of resources and tutorials for learning generative AI Events and Conferences Discussion about upcoming events and conferences related to generative AI Sharing of links to events such as the International Conference on Learning Representations (ICLR) and the Conference on Neural Information Processing Systems (NeurIPS) Social Media and Communities Sharing of links to social media groups and communities related to generative AI Discussion about the benefits of being part of a community for learning and networking Mention of the Generative Models Slack community and the Machine Learning subreddit Miscellaneous Discussion about the use of generative AI in video games Sharing of links to generative art websites such as Artbreeder and Deep Dream Generator Mention of the potential of generative AI in creating personalized content for individuals Links The description and link can be mismatched because of extraction errors.

AI Bot Code The group chat mentions that the Hasura AI bot code has been open-sourced and can be easily deployed with one click. They share a link to read more about it on https://theresanaiforthat.com/. AI Art The group chat discusses Offset Noise, which is apparently heating up AI art. They share a link to a tweet from MetaAI about it: https://twitter.com/MetaAI/status/1631351811696394240?s=20. Purchase Parity One member of the group mentions that the person behind Offset Noise is also running purchase parity, which they feel most products miss.

Generative AI: GoogleAI’s new Flan-UL2 20B is out Discussion on compute required for good/useful throughput Comparison with chatGPT API in terms of cost and data localization/compliance needs Need for specialized models and configurable/finetunable base models Two interesting articles from a16z on generative AI platform and everyday AI for consumers Links to personal articles on generative AI product strategy and AI products arms race Discussion on fMRI to SD research from Japan and its limitations Personal experience with off-the-shelf brain devices and noise in data Use of fMRI data for cleaner and more interesting analyses Request for resources on using SD for designers and artists Recommendation for Automatic1111 for using SD with plugins, but requires GPU and technical setup Excitement over possibility of getting actual footage of dreams Miscellaneous:

ChatGPT API Launch and Updates OpenAI provided an update on ChatGPT and Whisper API. A Twitter user shared a link to the ChatGPT API launch. A user mentioned getting controlnet working in a satisfactory way. Someone asked about deploying gpt-3.5-turbo on production apps and how it performs at scale. Another user replied that it is mostly stable but may require retries with exponential backoff. Corridor Digital’s Anime Creation StabilityAI shared a link to a tweet about Corridor Digital’s process for creating anime using SD VFX and Blender.

Introductions Members introduce themselves and their backgrounds in AI, ML, and related fields. Projects and Startups Pranjal Mehta talks about his electric plane startup, ePlane.ai, and his interest in building in GenAI. Nirant K works with startups building in ML. Kush is running Corp Dev for Glance & InMobi and is interested in seeing Generative AI innovations in the market. Pranav leads Peppertype.ai and has been closely working with Gen AI and its applications in content and marketing for the last 2 years.