This Day in AI Podcast

Show Notes: https://thisdayinai.com/bookmarks/28-ep51/
Sign up for daily This Day in AI: https://thisdayinai.com
Try Stable Cascade: https://simtheory.ai/agent/508-stable-cascade
Join SimTheory: https://simtheory.ai
======

This week we take several shots of vodka before trying to make sense of all the announcements. OpenAI attempted to trump Google's Gemini 1.5 with the announcement of Sora, 1 minute video generation that does an incredible job of keeping track of objects. Google showed us that up to 10M context windows are possible with multi-modal inputs. We discuss if a larger context window could end the need for RAG and take a first look at GraphRAG by Microsoft hoping to improve RAG with a knowledge graph. We road test Nvidia's ChatRTX on our baller graphics cards and Chris tries to delete all of his files using Microsoft UFO, a new open source project that uses GPT-4 vision to navigate and execute tasks on your Windows PC. We cover briefly V-JEPA (will try for next weeks show) and it's ability to learn through watching videos and listening, and finally discuss Stability's Stable Cascade which we've made available for "research" on SimTheory.

If you like the show please consider subscribing and leaving a comment. We appreciate your support.

======
Chapters:
00:00 - OpenAI's Sora That Creates Videos Instantly From Text
13:49 - ChatGPT Memory Released in Limited Preview
23:31 - OpenAI Rumored To Be Building Web Search, Andrej Karpathy Leaves OpenAI, Have OpenAI Slowed Down?
33:04 - Google Announces Gemini Pro 1.5. Huge Breakthrough 10M Context Window!
50:11 - Microsoft Research Publishes GraphRAG: Knowledge Graph Based RAG
1:02:03 - Nvidia's ChatRTX Road Tested
1:07:18 - AI Computers, AI PCs & Microsoft's UFO: An Agent for Window OS Interaction. Risk of AI Computers.
1:18:46 - Meta's V-JEPA: new architecture for self-supervised learning
1:24:26 - Stability AI's Stable Cascade

Subscribe to ThisDayInAI: https://thisdayinai.com
Try AI Agents on SimTheory: https://simtheory.ai
Show notes: https://thisdayinai.com/bookmarks/6-ep50

Tell us your thoughts on Gemini here: https://thisdayinai.com/post/62-your-thoughts-gemini-advanced/

Thanks to everyone for all your support and kind reviews to reach 50 episodes! Please consider leaving us a review wherever you get your podcasts.
=====

This week we cover the launch of Google Gemini Advanced, Gemini Ultra 1.0 and Bard being Renamed to Gemini. We compare GPT-4, Gemini Ultra 1.0 and Qwen 1.5 72B by sports betting $1000 on horse racing.

We celebrate 50 episodes and share our excited for Qwen 1.5 72B's performance at coding and quick refusals. We cover new releases including SyncLabs and Retell AI and Apple's Open Source Guiding Instruction-based Image Editing via Multimodal Large Language Models.

Finally, we discuss GOODY-2 and it's high refusal rate.

=====
CHAPTERS:

00:00 - Betting $1,000 To Compare Gemini Ultra 1.0 to GPT-4 to Qwen 1.5
07:33 - Google Gemini Advanced, Ultra: Details of Announcement and First Impressions
25:48 - OpenAI is Developing Agents to Control Your Devices
27:40 - Celebrating 50 Episodes of This Day in AI
30:34 - Qwen 1.5 72B: We're Impressed!
42:47 - SyncLabs: Tested & Impressions
47:58 - Retell AI: Tested & Impressions
54:18 - Apple's Open Source Guiding Instruction-based Image Editing via Multimodal Large Language Models
58:10 - GOODY-2: The World's Most Responsible AI Model

Join our new community: https://thisdayinai.com.
View the show notes here: https://thisdayinai.com/bookmarks/2-ep49/
Build AI Agents & Try AI From The Show: https://simtheory.ai

If you enjoy the podcast, please consider leaving us a review wherever you get your podcasts.

====
In this episode we reveal the new ThisDayinAI.com community website. We discuss the latest GPT-4 updates, Code Llama 70B open-source release and first impressions, we play around with the new LLaVA-1.6 release and are impressed by its capabilities. We also look at YOLO World and discuss the impact of EAGLE-7B and RWKV Language Models. Finally, we cover Bard's horrible new image creation feature and censorship. 

CHAPTERS:
====
00:00 - Introducing ThisDayInAI.com Community
5:10 - Be Careful What You Wish For! Mike Gets Spam Called by AI
16:16 - OpenAI Announces "improved" GPT-4 Preview Model to Make GPT-4 Less Lazy
27:00 - LLaVA-1.6: Improved reasoning, OCR, and world knowledge
34:00 - YOLO-World: Real-Time Open-Vocabulary Object Detection
45:11 - RWKV an RNN with GPT-level LLM performance and EAGLE7B Impressions
58:16 - Google Bard's New Highly Censored Image Creation Feature
1:07:13 - Will Google Bard be Renamed to Google Gemini?

Thanks for listening, we appreciate your support of the podcast.

This week we discuss Mark Zuckerberg confirming Llama 3, road test Elevenlabs Voice Dubbing, the state of AI apps and subscriptions, practical use cases of AI interacting with our world, does RAG make ChatGPT worse? Prompt compression with LongLLMLingua and how it might solve the attention problem, experiments with new image models including PhotoMaker and some LOLs to end the show.

AGENTS MENTIONED ON SHOW:
======
AI Phone Call On SimTheory:
https://simtheory.ai/agent/332-flirtatious-phone-call-assistant
MidJourney 6 on SimTheory:
https://simtheory.ai/agent/395-midjourney-image-creator
MidJourney 6 Video Creator:
https://simtheory.ai/agent/400-animate-midjourney-images

MORE LINKS:
======
Join the Discord: https://discord.gg/3gxM9H8qpv
Build an AI Agent: https://simtheory.ai

To support the show (and if you enjoy it) please consider becoming a paying subscriber to SimTheory to help us cover costs of agents, models and experiments we do for the show. Plus get access to every model, modality and the latest AI tech e.g. phone calling in a single place.

CHAPTERS
======
00:00 - Mark Zuckerberg Confirmed Llama 2 In Training
03:39 - Elevenlabs Voice Dubbing Service Tested
09:28 - Discussion on Research Labs, Apps & Future of AI App Business Models
18:43 - Bland.ai Update with Real World Examples & The Future of AI Agents & Agency interacting with our "analogue world"
30:56 - Nick Dobos Says RAG Makes ChatGPT Worse. Can Compression Help?
35:32 - LongLLMLingua and Prompt Compression
46:45 - Image Models: Photo Maker & Experiments with Image Generation
1:01:45 - LOLs including Rabbit r1 Fail, Claude Multi-Modal Leak, DPD Chat

SOURCES
======
https://www.youtube.com/watch?v=YeemJlrNx2Q
https://twitter.com/Stocktwits/status/1748043532340789570
https://twitter.com/danhendrycks/status/1749316795138552228
https://elevenlabs.io/dubbing
https://twitter.com/natfriedman/status/1750199867308433634
https://twitter.com/NickADobos/status/1749837866300264529?s=20
https://twitter.com/NickADobos/status/1749957909449187837?s=20
https://lumiere-video.github.io/
https://twitter.com/felix_red_panda/status/1749522604027682946?s=20
https://twitter.com/andrewcurran_/status/1747661100865511750?s=46
https://twitter.com/ashbeauchamp/status/1748034519104450874?s=20

PAPERS
======
https://arxiv.org/pdf/2310.06839.pdf
https://arxiv.org/pdf/2310.06839.pdf
https://arxiv.org/pdf/2312.04461v1.pdf

Build AI Agents & Try AI Agents From The Show On SimTheory: https://simtheory.ai
Join Discord: https://discord.gg/aphwE5snuq
Get Merch: https://www.thisdayinaimerch.com/

DESCRIPTION
====
In this episode, we dive into the buzz around GPT-5, sparked by Sam Altman's revelations on Bill Gates' latest podcast. We share our top hopes and dreams for GPT-5 and future AI advancements. Next, we delve into Microsoft's new CoPilot Pro Subscription, exploring how it stands out from ChatGPT Plus. Chris takes AutoGen Studio for a spin and ponders over its ideal user base. The episode then shifts to the intriguing concept of collaborative AI agents - is this the path to AI's mastering reasoning, reflection, and profound thought? We dissect the insights from the SeeAct Web Agents study, assessing its influence on AI agent development. Shifting gears, we discuss Google AMIE's groundbreaking ability to outperform doctors in diagnoses, even those assisted by AI. To wrap up, we spotlight the significance of Anthropic's Sleeper Agents experiment and its groundbreaking findings.

Thanks for listening. Please consider subscribing if you haven't already and leaving a review. We appreciate all of your support!

CHAPTERS:
====
00:00 - Cold Open
00:31 - GTP-5 Rumors & Leaks
07:32 - Microsoft CoPilot Pro
22:27 - Microsoft's AutoGen Studio: An open-source UI for AutoGen
38:53 - The Future of AI Agents? LAMs and SeeACT Web Agent Paper
1:00:19 - Google AMIE: Can AI Replace Doctors for Diagnosis?
1:13:12 -Anthropic's Sleep Agents Experiment

SOURCES:
====
https://twitter.com/arrakis_ai/status/1745672203683942863?s=20
https://twitter.com/daniacostaai/status/1746554047878824409?s=46
https://blogs.microsoft.com/blog/2024/01/15/bringing-the-full-power-of-copilot-to-more-people-and-businesses/
https://twitter.com/emollick/status/1747359731595763817
https://microsoft.github.io/autogen/blog/2023/12/01/AutoGenStudio/
https://osu-nlp-group.github.io/SeeAct/
https://blog.research.google/2024/01/amie-research-ai-system-for-diagnostic_12.html
https://www.bloomberg.com/news/articles/2024-01-14/artificial-intelligence-will-affect-almost-40-of-jobs-imf-says
https://twitter.com/Teknium1/status/1746067427379798344

PAPERS:
====
https://arxiv.org/pdf/2401.01614.pdf
https://arxiv.org/pdf/2401.05654.pdf
https://arxiv.org/pdf/2401.05566.pdf

Try AI Voice Calling: https://simtheory.ai/agent/332-flirtatious-phone-call-assistant
Join Our Discord: https://discord.gg/s7bCFV4gTr
Join SimTheory: https://simtheory.ai

In this episode we put Bland.ai to the test. We try out their new AI technology for voice calls that can react and respond in near real time by prank calling our local hardware and pet stores.

We also discuss the launch of more AI dedicated hardware in the Rabit r1, the GPT Store now it's finally released with over 3M GPTs, discuss GPT Teams, LUMA, AudioBox and ask, are we in an AI bubble?

If you like this episode please consider liking, subscribing and commenting. Thanks for watching!

CHAPTERS
====
00:00 - Our call to the hardware store
00:30 - Bland.ai Voice Calling with AI
03:04 - Prank Calling a Hardware Store with AI
11:21 - Calling a Pet Grooming Store with AI
18:15 - Thoughts in AI Hardware, Cherry Picked AI Demos & Rabit r1
35:35 - OpenAI Releases GPT Store with 3M GPTs, Cloning Problem & Initial Reactions
45:22 - OpenAI Releases ChatGPT Teams
47:57 - ChatGPT Memory
49:26 - LUMA Genie, The Metaverse & Vision Pro Apps
55:05 - The AI Jailbreak Problem & Rethinking Persuasion to Challenge AI Safety by Humanizing LLMs
1:00:52 - Meta AudioBox
1:03:38 - Microsoft Overtakes Apple as Most Valuable Company - Is it because of AI? And is AI a Bubble?

SOURCES:
====
https://twitter.com/usebland/status/1743411488612913429
https://chats-lab.github.io/persuasive_jailbreaker/
https://www.rabbit.tech/
https://twitter.com/Dan_Jeffries1/status/1745404485298459106
https://twitter.com/abacaj/status/1745474794638745892?s=20
https://twitter.com/AravSrinivas/status/1745489529551905159
https://openai.com/blog/introducing-the-gpt-store
https://twitter.com/NickADobos/status/1745244031381291164/photo/2
https://twitter.com/AndrewCurran_/status/1744918982174429432?s=20
https://openai.com/blog/introducing-chatgpt-team
https://lumalabs.ai/genie
https://audiobox.metademolab.com/maker
https://www.miri.health/

It's great to be back! In this episode we cover everything new and everything we missed during our break. We start with breaking news that the OpenAI ChatGPT GPT Store is being released next week, then cover Gemini Pro and Gemini Pro Vision API, Mixtral APIs, AnyText, NY Times Copyright lawsuit and finally.. get excited about a dishwashing robot!

====
Join SimTheory: https://simtheory.ai
Join Discord: https://discord.gg/aphwE5snuq
Get Merch: https://www.thisdayinaimerch.com/

Try models from the show:
====
Gemini Pro: https://simtheory.ai/agent/282-google-gemini-assistant
Mixtral: https://simtheory.ai/agent/129-miss-mistra-mistral-medium
Stable Diffusion Video: https://simtheory.ai/agent/224-image-to-video-creation-agent
AI Movie Trailer Maker: https://simtheory.ai/agent/279-ai-movie-trailer-maker

CHAPTERS:
====
00:00 - Mike's AI Movie Trailer Intro
02:05 - GPT Store Will go Live Next Week
22:52 - Gemini Pro API & Gemini Pro Vision Road Tested (literally)
33:34 - Mixtral API: Mistral Platform API Tested
45:31 - Stable Video Diffusion
48:12 - Pika AI Video General Availability
52:05 - Stability AI Memberships
55:54 - Prompt Injection for DALL-E with Public Domain
57:34 - New York Times Sues OpenAI & Microsoft for Copyright Infringement
1:04:49 - Inpainting with AnyText
1:14:15 - Microsoft CoPilot App with GPT-4 Now On iOS and Android
1:14:39 - One More Thing: The Dishwasher Bot

SOURCES:
====
https://time.com/6551496/mickey-mouse-public-domain-steamboat-willie/
https://twitter.com/digthatdata/status/1742074049260621976?s=46
https://www.theguardian.com/media/2023/dec/27/new-york-times-openai-microsoft-lawsuit
https://www.reuters.com/technology/apple-explores-ai-deals-with-news-publishers-new-york-times-2023-12-22/
https://twitter.com/rowancheung/status/1742967393310368222/photo/1
https://www.theinformation.com/briefings/openai-to-launch-chatbot-store-next-week?rc=kvsmhw
https://blog.google/technology/ai/gemini-api-developers-cloud/
https://mistral.ai/news/mixtral-of-experts/
https://mistral.ai/news/la-plateforme/
https://stability.ai/news/stable-video-diffusion-open-ai-video-model
https://simtheory.ai/share/d49c8c00-9fda-40aa-b386-a7c27455015b/
https://pika.art/
https://stability.ai/membership
https://twitter.com/venturetwins/status/1742976476432196100?s=46
https://github.com/tyxsspa/anytext
https://www.theverge.com/2023/12/29/24019288/microsoft-copilot-app-available-iphone-ipad-ai
https://mobile-aloha.github.io/

Join our discord: https://discord.gg/zqz5fVyx7m
Get the merch: https://thisdayinaimerch.com
Try Agents & Models on SimTheory: https://simtheory.ai

In our final episode for the year, we cover the surprise announcement of Google's Gemini AI models and give our first impressions. We road test Gemini Pro on Bard and discuss the likely impact of Gemini on the market and developer ecosystems. Then it's time for our holiday gift: SimTheory. Now you can use AI agents we mention on the show including our virtual girlfriends, Sports Betting with AI and many more! You can even create your own agents to try different models using the same tools we use to prepare for the show. We then discuss if Ilya is OK and the drama at OpenAI. And finally, we make predictions for 2024 and cover some of Meta's latest announcements.

Thanks for watching, listening and all your support through 2023. We really appreciate it and will see you early next year!

CHAPTERS:
=====
00:00 - Google Gemini is Here? Kinda
38:48 - Our Holiday Gift: SimTheory: Virtual Girlfriend, Sports Betting with AI Agents
51:15 - Is Ilya OK? Is GPT-4 Slowness About Cost Reductions?
56:26 - NexusRaven-V2-13B for function calling: is this the future of specialized fine tune models?
1:00:14 - Our Predictions for AI in 2024
1:12:54 - Meta announces AI Alliance for AI Openness + Updates to Meta AI Characters and SeamlessExpressive
1:15:43 - Final thoughts and thank you

SOURCES:
=====
https://blog.google/technology/ai/google-gemini-ai/
https://twitter.com/tunguz/status/1732444203437695387
https://twitter.com/tunguz/status/1732444203437695387
https://twitter.com/tunguz/status/1732444203437695387
https://twitter.com/tunguz/status/1732444203437695387
https://techcrunch.com/2023/12/07/early-impressions-of-googles-gemini-arent-great/
https://twitter.com/clementdelangue/status/1732138699901809042
https://huggingface.co/Nexusflow/NexusRaven-V2-13B
https://twitter.com/abemurray/status/1732723510810759369
https://ai.meta.com/blog/ai-alliance
https://techcrunch.com/2023/12/06/metas-ai-characters-are-now-live-across-its-u-s-apps-with-support-for-bing-search-and-better-memory/
https://techcrunch.com/2023/12/06/meta-ai-adds-reels-support-and-reimagine-a-way-to-generate-new-ai-images-in-group-chats/
https://seamless.metademolab.com/expressive
https://twitter.com/mattrickard/status/1731889331516936261

Join the discord: https://discord.gg/27mQ9cut
Get the merch: https://thisdayinaimerch.com

This week we celebrate ChatGPT's 1 Year Anniversary and Ask is GPT-4 Lazy? We explore the best of open source with Wizard 33B and test China's Qwen 72B model from Alibaba. Chris tries to delete all files from his computer using Self Operation AI Computer and we cover Amazon's AWS Ignite AI announcements, Stability Diffusion XL Turbo, The Scalable Extraction Attack on ChatGPT and an exciting waitlist release from PIKA. 

Like, sub, comment if you enjoy the episode to support the show. We love hearing from you.

CHAPTERS:
=====
00:00 - Cold Open
00:08 - ChatGPT 1 Year Anniversary
07:54 - Is GPT-4 Lazy? Is Claude Unusable Now?   
18:43 - Are Open-Source Models Catching Up 1 Year On?
21:57 - Wizard 33B Open-Source Model
24:55 - Demo of Wizard 33B
28:26 - China's Qwen 72B Open-Source Model
31:26 - Qwen Demo
38:16 - Self Operation Computer Discussion & The Future of AI With Access to Computers
49:23 - Scalable Extraction: DeepMind's COMPANY attack to extract training data from ChatGPT
55:20 - Stability Diffusion XL Turbo, Stability's Stability & Commercial Subscriptions
1:03:23 - Amazon's AWS Ignite: Amazon Q, Trainium 2, Bedrock Fine Tuning
1:07:49 - PIKA Video
1:09:26 - Important News

SOURCES:
======
https://arstechnica.com/information-technology/2023/11/chatgpt-was-the-spark-that-lit-the-fire-under-generative-ai-one-year-ago-today/
https://twitter.com/emollick/status/1729604442826170586?s=46
https://twitter.com/krishnanrohit/status/1729353613498261597?s=46
https://arxiv.org/pdf/2311.16989.pdf
https://twitter.com/huybery/status/1730127387109781932/photo/1
https://arxiv.org/pdf/2309.16609.pdf
https://github.com/OthersideAI/self-operating-computer/tree/main
https://arxiv.org/pdf/2311.17035.pdf
https://twitter.com/ayushsoni_io/status/1730128497572462695
https://stability.ai/news/stability-ai-sdxl-turbo
https://pika.art/
https://venturebeat.com/ai/amazon-awss-barrage-of-gen-ai-announcements-aim-to-outdo-microsoft/

Join Our Discord: https://discord.gg/58HtZnVD
Buy The Merch: https://www.thisdayinaimerch.com/

This week we reluctantly cover all the OpenAI drama and ask What Did Sam Altman Actually Do? Is Q* a path to AGI or just one big "look over here" distraction so we stop asking all these questions... We also cover Andrej Karpathy's LLM OS vision, discuss Claude 2.1 and how bad it's become thanks to "safety" and discuss our initial impressions of Stable Video Diffusion. Finally, we have some fun with Suno!

If you like this podcast, please consider subscribing and liking this episode. We appreciate the support.

CHAPTERS:
====
00:00 - A Full Recap of What Happened with Sam Altman & OpenAI
10:06 - What Did Sam Altman Actually Do? 
28:03 - What Did Ilya Really Discover? Is Q* A Big Distraction? How Far Ahead if OpenAI?
40:47 - Will This Drama Help Progress Open Source AI?
51:11 - Is Andrej Karpathy's LLM OS Vision The Future?
1:00:25 - Inflection-2 LLM
1:02:35 - Stable Video Diffusion Initial Thoughts
1:06:40 - Claude 2.1 Announcement 200K Context
1:21:26 Fun with Suno AI: Make Music with a Prompt

SOURCES:
====
https://www.theinformation.com/briefings/openais-employee-share-sale-to-continue-after-altman-returns?rc=kvsmhw
https://twitter.com/openai/status/1727236805182026159?s=46
https://openai.com/blog/openai-announces-leadership-transition
https://twitter.com/ylecun/status/1727727656118923296?s=46
https://www.theinformation.com/articles/openai-dramas-first-season-ends-but-second-season-is-possible?rc=kvsmhw
https://www.theinformation.com/articles/openai-made-an-ai-breakthrough-before-altman-firing-stoking-excitement-and-concern?rc=kvsmhw
https://inflection.ai/inflection-2
https://stability.ai/news/stable-video-diffusion-open-ai-video-model
https://www.anthropic.com/index/claude-2-1
https://app.suno.ai/
https://www.reuters.com/technology/sam-altmans-ouster-openai-was-precipitated-by-letter-board-about-ai-breakthrough-2023-11-22/