Episodes

Tuesday Feb 06, 2024
Tuesday Feb 06, 2024
Colossyan uses GenAI to create engaging corporate training videos. Learn how to train your own Generative Adversarial Network (GAN). Researchers propose a new method for LLMs to better leverage tools in multi-step reasoning. China's tech giants heat up the generative video race. Explore the latest advancements in AI technology and their potential impact on various industries.Sources:https://techcrunch.com/2024/02/06/colossyan-uses-genai-to-create-corporate-training-videos/https://www.analyticsinsight.net/how-to-train-your-own-generative-adversarial-network/https://www.marktechpost.com/2024/02/05/researchers-from-epfl-and-meta-ai-proposes-chain-of-abstraction-coa-a-new-method-for-llms-to-better-leverage-tools-in-multi-step-reasoning/https://techcrunch.com/2024/02/05/chinas-generative-video-race-heats-up/Outline:(00:00:00) Introduction(00:00:44) Colossyan uses GenAI to create corporate training videos(00:03:15) How to Train Your Own Generative Adversarial Network(00:05:21) Researchers from EPFL and Meta AI Proposes Chain-of-Abstraction (CoA): A New Method for LLMs to Better Leverage Tools in Multi-Step Reasoning(00:08:15) China’s generative video race heats up

Monday Feb 05, 2024
Monday Feb 05, 2024
7 new AI tools beat ChatGpt again, showcasing the remarkable advancements in AI generation technology. Felicis, a renowned venture firm, shares insights on assessing and attracting winners in a crowded AI startup landscape. Plus, groundbreaking research unveils WRAP, a game-changer for pre-training language models with synthetic data. Microsoft researchers introduce StrokeNUWA, tokenizing strokes for vector graphic synthesis.Sources:https://heavenvalley.medium.com/7-new-ai-tools-beat-chatgpt-again-9836210b00cbhttps://techcrunch.com/2024/02/04/how-vcs-can-assess-and-attract-winners-in-a-landscape-thats-now-crowded-with-ai-startups/https://www.marktechpost.com/2024/02/04/this-ai-paper-from-cmu-and-apple-unveils-wrap-a-game-changer-for-pre-training-language-models-with-synthetic-data/https://www.marktechpost.com/2024/02/04/microsoft-researchers-introduce-strokenuwa-tokenizing-strokes-for-vector-graphic-synthesis/Outline:(00:00:00) Introduction(00:00:40) 7 New Ai Tools Beat ChatGpt Again(00:03:21) How VCs can assess and attract winners in a landscape that’s now crowded with AI startups(00:06:13) This AI Paper from CMU and Apple Unveils WRAP: A Game-Changer for Pre-training Language Models with Synthetic Data(00:08:48) Microsoft Researchers Introduce StrokeNUWA: Tokenizing Strokes for Vector Graphic Synthesis

Sunday Feb 04, 2024
Sunday Feb 04, 2024
Discover the groundbreaking Eagle 7B AI model with 7.52 billion parameters and its potential to revolutionize industries while minimizing environmental impact. Learn how Microsoft's profit soared 33% on AI and cloud-computing investments, positioning the company as a frontrunner in the AI race. Join us as we explore the latest advancements in AI and their implications for various sectors.Sources:https://www.marktechpost.com/2024/02/03/meet-eagle-7b-a-7-52b-parameter-ai-model-built-on-the-rwkv-v5-architecture-and-trained-on-1-1t-tokens-across-100-languages/https://www.marktechpost.com/2024/02/03/this-ai-paper-from-china-introduces-segmamba-a-novel-3d-medical-image-segmentation-mamba-model-designed-to-effectively-capture-long-range-dependencies-within-whole-volume-features-at-every-scale/https://www.marktechpost.com/2024/02/03/researchers-from-the-university-of-washington-developed-a-deep-learning-method-for-protein-sequence-design-that-explicitly-models-the-full-non-protein-atomic-context/https://apnews.com/article/microsoft-earnings-results-9a81e475deef2baec8cdfdb58142d475Outline:(00:00:00) Introduction(00:01:00) Meet Eagle 7B: A 7.52B Parameter AI Model Built on the RWKV-v5 architecture and Trained on 1.1T Tokens Across 100+ Languages(00:03:33) This AI Paper from China Introduces SegMamba: A Novel 3D Medical Image Segmentation Mamba Model Designed to Effectively Capture Long-Range Dependencies within Whole Volume Features at Every Scale(00:05:52) Researchers from the University of Washington Developed a Deep Learning Method for Protein Sequence Design that Explicitly Models the Full Non-Protein Atomic Context(00:08:47) Microsoft profit soars 33% on AI, cloud-computing investments

Saturday Feb 03, 2024
Saturday Feb 03, 2024
Discover groundbreaking research on the connection between AI and the human brain. Learn about Dify.AI, an open-source platform ensuring privacy and customization in AI. Explore SliceGPT, a revolutionary sparsification scheme for efficient compression of language models. Delve into the impacts of AI deepfakes and the need for awareness and mitigation. Join us as we explore the latest advancements and challenges in the world of AI.Sources:https://www.marktechpost.com/2024/02/02/unlocking-the-brains-language-response-how-gpt-models-predict-and-influence-neural-activity/https://www.marktechpost.com/2024/02/02/meet-dify-ai-an-llm-application-development-platform-that-integrates-baas-and-llmops/https://www.marktechpost.com/2024/02/02/researchers-from-eth-zurich-and-microsoft-introduce-slicegpt-for-efficient-compression-of-large-language-models-through-sparsification/https://technology.inquirer.net/131679/impact-of-deepfakeOutline:(00:00:00) Introduction(00:00:49) Unlocking the Brain’s Language Response: How GPT Models Predict and Influence Neural Activity(00:03:40) Meet Dify.AI: An LLM Application Development Platform that Integrates BaaS and LLMOps(00:06:22) Researchers from ETH Zurich and Microsoft Introduce SliceGPT for Efficient Compression of Large Language Models through Sparsification(00:09:15) AI deepfakes and their deeper impacts

Friday Feb 02, 2024
Friday Feb 02, 2024
Bank of England chief dismisses AI job destruction fears. Introducing FUSIONL AI, a startup revolutionizing learning with its Smart Minimalistic Language Model. Plus, UK ministers urged to protect creatives from unauthorized AI use and antitrust enforcers race to understand how to tackle AI.Sources:https://www.theguardian.com/business/live/2024/feb/02/ai-jobs-bank-of-england-goldrush-us-jobs-report-business-livehttps://medium.com/@unravelingentertainment/introducing-fusionl-ai-revolutionizing-learning-with-smart-minimalistic-language-model-2b9f6cffc649https://www.theguardian.com/technology/2024/feb/02/uk-ministers-urged-to-protect-creatives-whose-work-is-used-by-ai-firmshttps://techcrunch.com/2024/02/01/eu-antitrust-new-world-order/Outline:(00:00:00) Introduction(00:00:42) AI will not be mass destroyer of jobs, says Bank of England chief(00:03:37) Introducing FUSIONL AI: Revolutionizing Learning with Smart Minimalistic Language Model(00:06:13) UK ministers urged to protect creatives whose work is used by AI firms(00:09:40) Antitrust enforcers admit they’re in a race to understand how to tackle AI

Thursday Feb 01, 2024
Thursday Feb 01, 2024
Deciphering neuronal universality in GPT-2 language models. Revolutionizing orchard management with deep learning. The vulnerabilities of vision-language models and the first red teaming dataset. Neural networks on the brink of universal prediction with DeepMind's cutting-edge approach.Sources:https://www.marktechpost.com/2024/01/31/deciphering-neuronal-universality-in-gpt-2-language-models/https://www.spectroscopyonline.com/view/revolutionizing-orchard-management-deep-learning-yields-precise-fruit-tree-segmentationhttps://www.marktechpost.com/2024/01/31/this-ai-paper-from-china-sheds-light-on-the-vulnerabilities-of-vision-language-models-unveiling-rtvlm-the-first-red-teaming-dataset-for-multimodal-ai-security/https://syncedreview.com/2024/01/31/neural-networks-on-the-brink-of-universal-prediction-with-deepminds-cutting-edge-approach/Outline:(00:00:00) Introduction(00:00:45) Deciphering Neuronal Universality in GPT-2 Language Models(00:03:17) Revolutionizing Orchard Management: Deep Learning Yields Precise Fruit Tree Segmentation(00:06:42) This AI Paper from China Sheds Light on the Vulnerabilities of Vision-Language Models: Unveiling RTVLM, the First Red Teaming Dataset for Multimodal AI Security(00:08:56) Neural Networks on the Brink of Universal Prediction with DeepMind’s Cutting-Edge Approach

Wednesday Jan 31, 2024
Wednesday Jan 31, 2024
This AI paper unveils the future of MultiModal Large Language Models (MM-LLMs) and their impact on AI research. Meet Spade, an AI method for automatically synthesizing assertions that identify bad LLM outputs. Plus, AI and machine learning to keep old structures strong, and researchers introduce CoEdIT, an AI-based text editing system designed to provide writing assistance with a natural language interface.Sources:https://www.marktechpost.com/2024/01/30/this-ai-paper-unveils-the-future-of-multimodal-large-language-models-mm-llms-understanding-their-evolution-capabilities-and-impact-on-ai-research/https://www.marktechpost.com/2024/01/30/meet-spade-an-ai-method-for-automatically-synthesizing-assertions-that-identify-bad-llm-outputs/https://www.newindianexpress.com/xplore/2024/Jan/31/ai-machine-learning-to-keep-old-structures-stronghttps://www.marktechpost.com/2024/01/30/researchers-from-grammarly-and-the-university-of-minnesota-introduce-coedit-an-ai-based-text-editing-system-designed-to-provide-writing-assistance-with-a-natural-language-interface/

Tuesday Jan 30, 2024
Tuesday Jan 30, 2024
A groundbreaking compact large vision language model for standard GPUs is unveiled in China. Microsoft's acquisition of Activision and its bets on AI yield high quarterly revenue. The Taylor Swift deepfake debacle highlights content moderation failures and the need for a complete overhaul of social platforms. Stay informed about the latest advancements in AI technology and the challenges it presents.Sources:https://www.marktechpost.com/2024/01/30/this-ai-paper-from-china-unveils-vary-toy-a-groundbreaking-compact-large-vision-language-model-for-standard-gpus-with-advanced-vision-vocabulary/https://www.theguardian.com/technology/2024/jan/30/microsoft-revenue-activision-ai-surpassed-earningshttps://apnews.com/article/microsoft-earnings-results-9a81e475deef2baec8cdfdb58142d475https://techcrunch.com/2024/01/30/the-taylor-swift-deepfake-debacle-was-frustratingly-preventable/Outline:(00:00:00) Introduction(00:00:45) This AI Paper from China Unveils ‘Vary-toy’: A Groundbreaking Compact Large Vision Language Model for Standard GPUs with Advanced Vision Vocabulary(00:03:32) Microsoft’s Activision acquisition and bets on AI yield high quarterly revenue(00:06:31) Microsoft profit soars 33% on AI, cloud-computing investments(00:09:28) The Taylor Swift deepfake debacle was frustratingly preventable

Monday Jan 29, 2024
Monday Jan 29, 2024
AI companies will need to report safety tests to the US government. New hope for early pancreatic cancer intervention via AI-based risk prediction. OpenAI faces allegations of violating Europe's privacy laws. Stay informed about the latest developments in AI regulation, healthcare advancements, and privacy concerns.Sources:https://apnews.com/article/biden-ai-artificial-intelligence-safe-395591bcde523416db88767fa54f30f5https://news.mit.edu/2024/new-hope-early-pancreatic-cancer-intervention-ai-based-risk-prediction-0118https://techcrunch.com/2024/01/29/chatgpt-italy-gdpr-notification/https://apnews.com/article/biden-ai-artificial-intelligence-safe-395591bcde523416db88767fa54f30f5

Friday Jan 26, 2024
Friday Jan 26, 2024
The rapid online spread of deepfake pornographic images featuring Taylor Swift has sparked renewed calls for US legislation. Researchers at MIT have developed AI models for early detection of pancreatic cancer. Plus, MIT PhD students are revolutionizing natural language models for improved reliability. Also, OpenAI drops prices and fixes 'lazy' GPT-4 model.Sources:https://www.theguardian.com/music/2024/jan/26/taylor-swift-deepfake-pornography-sparks-renewed-calls-for-us-legislationhttps://news.mit.edu/2024/new-hope-early-pancreatic-cancer-intervention-ai-based-risk-prediction-0118https://news.mit.edu/2024/reasoning-and-reliability-in-ai-0118https://techcrunch.com/2024/01/25/openai-drops-prices-and-fixes-lazy-gpt-4-that-refused-to-work/