Last Week in AI-logo

Last Week in AI

Technology Podcasts

Weekly summaries and discussion about the most interesting developments in AI, deep learning, robotics, and more!

Location:

United States

Description:

Weekly summaries and discussion about the most interesting developments in AI, deep learning, robotics, and more!

Language:

English


Episodes
Ask host to enable sharing for playback control

#170 - new Sora rival, OpenAI robotics, understanding GPT4, AGI by 2027?

6/9/2024
Our 170th episode with a summary and discussion of last week's big AI news! With hosts Andrey Kurenkov (https://twitter.com/andrey_kurenkov) and Jeremie Harris (https://twitter.com/jeremiecharris) Feel free to leave us feedback here. Read out our text newsletter and comment on the podcast at https://lastweekin.ai/ Email us your questions and feedback at contact@lastweekin.ai and/or hello@gladstone.ai Timestamps + Links: KLING is the latest AI video generator that could rival OpenAI's Sora‘Apple Intelligence’ will automatically choose between on-device and cloud-powered AIUdio introduces new udio-130 music generation model and more advanced featuresPerplexity AI’s new feature will turn your searches into shareable pagesElevenLabs’ AI generator makes explosions or other sound effects with just a promptGoogle’s updated AI-powered NotebookLM expands to India, UK and over 200 other countriesOpenAI is restarting its robotics research groupSaudi fund invests in China effort to create rival to OpenAIUAE seeks ‘marriage’ with US over artificial intelligence dealsZoox to test self-driving cars in Austin and Miami Microsoft Lays Off 1,500 Workers, Blames "AI Wave"Avengers, assemble—Google, Intel, Microsoft, AMD and more team up to develop an interconnect standard to rival Nvidia's NVLinkGLM-4-9B-Chat-1MHugging Face and Pollen Robotics show off first project: an open source robot that does choresZyphra debuts Zyda, a 1.3T language modeling dataset it claims outperforms Pile, C4, arxivStability AI debuts new Stable Audio Open for sound designScaling and evaluating sparse autoencodersImproving Alignment and Robustness with Short CircuitingAutomatic Data Curation for Self-Supervised Learning: A Clustering-Based ApproachGPT-4 didn't ace the bar exam after all, MIT research suggests — it didn't even break the 70th percentileFormer OpenAI researcher foresees AGI reality in 2027OpenAI Insiders Warn of a ‘Reckless’ Race for DominanceTesting and mitigating elections-related risksTeams of LLM Agents can Exploit Zero-Day VulnerabilitiesThe Uncanny Rise of the World's First AI Beauty Pageant

Duration:01:48:30

Ask host to enable sharing for playback control

#169 - Google's Search Errors, OpenAI news & DRAMA, new leaderboards

6/2/2024
Our 168th episode with a summary and discussion of last week's big AI news! Feel free to leave us feedback here: https://forms.gle/ngXvXZpNJxaAprDv6 Read out our text newsletter and comment on the podcast at https://lastweekin.ai/ Email us your questions and feedback at contact@lastweekin.ai and/or hello@gladstone.ai Timestamps + Links: Google’s A.I. Search Errors Cause a Furor OnlineTelegram gets an in-app Copilot botOpera is adding Google's Gemini AI to its browserAmazon plans to give Alexa an AI overhaul — and a monthly subscription priceMicrosoft Edge will translate and dub YouTube videos as you’re watching themIyo thinks its gen AI earbuds can succeed where Humane and Rabbit stumbledPwC agrees deal to become OpenAI's first reseller and largest enterprise userVox Media and The Atlantic sign content deals with OpenAIOpenAI launches programs making ChatGPT cheaper for schools and nonprofitsHuawei patent reveals 3nm-class process technology plans — China continues to move forward despite US sanctionsNvidia, Powered by A.I. Boom, Reports Soaring Revenue and ProfitsElon Musk’s xAI raises $6 billion in latest funding roundScale AI publishes its first LLM Leaderboards, ranking AI model performance in specific domainsCohere For AI Launches Aya 23, 8 and 35 Billion Parameter Open Weights ReleaseWho will make AlphaFold3 open source? Scientists race to crack AI modelMistral releases Codestral, its first generative AI model for codeThe Road Less ScheduledTraining Compute of Frontier AI Models Grows by 4-5x per Yeargzip Predicts Data-dependent Scaling LawsNeural Scaling Laws for Embodied AIContextual Position Encoding: Learning to Count What’s ImportantNew AI products much hyped but not much used, study saysEx-OpenAI board member reveals what led to Sam Altman's brief oustingOpenAI researcher who resigned over safety concerns joins AnthropicLeaked OpenAI Documents Show Sam Altman Was Clearly Aware of Silencing Former EmployeesOpenAI Board Forms Safety and Security CommitteeRobocaller Who Used AI to Clone Biden’s Voice Fined $6 MillionHacker Releases Jailbroken "Godmode" Version of ChatGPTChina Creates $47.5 Billion Chip Fund to Back Nation’s FirmsAlphabet, Meta Offer Millions to Partner With Hollywood on AI

Duration:02:06:22

Ask host to enable sharing for playback control

#168 - OpenAI vs Scar Jo + safety researchers, MS AI updates, cool Anthropic research

5/28/2024
Our 168th episode with a summary and discussion of last week's big AI news! With guest host Gavin Purcell from AI for Humans podcast! Read out our text newsletter and comment on the podcast at https://lastweekin.ai/ Email us your questions and feedback at contact@lastweekin.ai and/or hello@gladstone.ai Timestamps + Links: OpenAI says Sky voice in ChatGPT will be paused after concerns it sounds too much like Scarlett JohanssonMicrosoft’s Copilot assistant is getting a GPT-4o upgradeRecall is Microsoft’s key to unlocking the future of PCsElevenLabs Launches AI-Voiced Screen Reader AppAdobe Lightroom gets a magic eraser, and it’s impressiveMicrosoft, Khan Academy provide free AI assistant for all educators in USMicrosoft Paint is getting an AI-powered image generator that responds to your text prompts and doodlesOpenAI founders Sam Altman and Greg Brockman go on the defensive after top safety researchers quitOpenAI, WSJ Owner News Corp Strike Content Deal Valued at Over $250 MillionCoreWeave Raises $7.5 Billion in Debt for AI Computing PushGoogle announced Trillium, its sixth generation of Tensor processors.Inflection AI reveals new team and plan to embed emotional AI in business botsData-labeling startup Scale AI raises $1B as valuation doubles to $13.8BAbacus AI Releases Smaug-Llama-3-70B-Instruct: The New Benchmark in Open-Source Conversational AI Rivaling GPT-4 TurboIntroducing New Chatbot Arena Category: Hard PromptsMicrosoft brings out a small language model that can look at picturesNew Anthropic Research Sheds Light on AI's 'Black Box'Chameleon: Mixed-Modal Early-Fusion Foundation ModelsSpeechVerse: A Large-scale Generalizable Audio Language ModelCAT3D: Create Anything in 3D with Multi-View Diffusion ModelsCoin3D: Controllable and Interactive 3D Assets Generation with Proxy-Guided ConditioningSpeechGuard: Exploring the Adversarial Robustness of Multimodal Large Language ModelsWorld’s first major law for artificial intelligence gets final EU green lightColorado governor signs sweeping AI regulation billSenators Propose $32 Billion in Annual A.I. Spending but Defer RegulationGoogle DeepMind launches new framework to assess the dangers of AI modelsTech giants pledge AI safety commitments — including a ‘kill switch’ if they can’t mitigate risksSony Music warns tech companies over ‘unauthorized’ use of its content to train AIHollywood agency CAA aims to help stars manage their own AI likenessesWhat Do You Do When A.I. Takes Your Voice?

Duration:01:44:31

Ask host to enable sharing for playback control

#167 - GPT-4o, Project Astra, Veo, OpenAI Departures, Interview with Andrey

5/19/2024
Our 167th episode with a summary and discussion of last week's big AI news! With guest host Daliana Liu (https://www.linkedin.com/in/dalianaliu/) from The Data Scientist Show! And a special one-time interview with Andrey in the latter part of the podcast. Read out our text newsletter and comment on the podcast at https://lastweekin.ai/ Email us your questions and feedback at contact@lastweekin.ai and/or hello@gladstone.ai Timestamps + links: OpenAI releases GPT-4o, a faster model that’s free for all ChatGPT usersProject Astra is the future of AI at GoogleGoogle is redesigning its search engine — and it’s AI all the way downGoogle unveils Veo and Imagen 3, its latest AI media creation modelsGoogle Unveils Music AI Sandbox Making Loops From PromptsAnthropic AI Launches a Prompt Engineering Tool that Generates Production-Ready Prompts in the Anthropic ConsoleOpenAI’s Chief Scientist and Co-Founder Is Leaving the CompanyMike Krieger joins Anthropic as Chief Product Officer$16k G1 humanoid rises up to smash nuts, twist and twirlGM's Cruise to start testing robotaxis in Phoenix area with human safety drivers on boardUS agency probes Amazon-owned Zoox self-driving vehicles after two crashesWaymo’s robotaxis under investigation after crashes and traffic mishapsIntroducing PaliGemma, Gemma 2, and an Upgraded Responsible AI ToolkitFalcon 2: UAE’s Technology Innovation Institute Releases New AI Model Series, Outperforming Meta’s New Llama 3License to Call: Introducing Transformers Agents 2.0The Platonic Representation HypothesisSUTRA: Scalable Multilingual Language Model ArchitectureBipartisan Senate bill on AI security would bolster voluntary cyber reporting processesU.K. agency releases tools to test AI model safetyProtesters Are Fighting to Stop AI, but They’re Split on How to Do ItGoogle’s invisible AI watermark will help identify generative text and videoHow One Author Pushed the Limits of AI CopyrightStellaris gets an DLC about AI that features AI-created voices, director insists it's 'ethical' and 'we're pretty good at exploring dystopian sci-fi and don't want to end up there ourselves'At the AI Film Festival, humanity triumphed over tech

Duration:01:43:02

Ask host to enable sharing for playback control

#166 - new AI song generator, Microsoft's GPT4 efforts, AlphaFold3, xLSTM, OpenAI Model Spec

5/12/2024
Our 166th episode with a summary and discussion of last week's big AI news! Read out our text newsletter and comment on the podcast at https://lastweekin.ai/ Email us your questions and feedback at contact@lastweekin.ai and/or hello@gladstone.ai Timestamps + links: ElevenLabs previews music-generating AI modelMysterious “gpt2-chatbot” AI model appears suddenly, confuses expertsSoundHound AI and Perplexity Partner to Bring Online LLMs to Next Gen Voice Assistants Across Cars and IoT DevicesStability AI sows gen AI discord with Stable ArtisanApple Will Revamp Siri to Catch Up to Its Chatbot CompetitorsAlibaba rolls out latest version of its large language model to meet robust AI demandOpenAI and Stack Overflow partner to bring more technical knowledge into ChatGPTNew Microsoft AI model may challenge GPT-4 and Google GeminiWayve, an A.I. Start-Up for Autonomous Driving, Raises $1 BillionMotional delays commercial robotaxi plans amid restructuringThe rise of the Chinese AI unicorns doing battle with OpenAIPrometheus 2: An Open Source Language Model Specialized in Evaluating Other Language ModelsDeepSeek-V2: A Strong, Economical, and Efficient Mixture-of-Experts Language ModelOpenVoice V2: Evolving Multilingual Voice Cloning with Enhanced Style Control and Cross-Lingual CapabilitiesGranite Code Models: A Family of Open Foundation Models for Code IntelligenceHugging Face launches LeRobot open source robotics code libraryVibe-Eval: A new open and hard evaluation suite for measuring progress of multimodal language modelsGoogle DeepMind’s Groundbreaking AI for Protein Structure Can Now Model DNAxLSTM: Extended Long Short-Term MemoryStoryDiffusion: Consistent Self-Attention for Long-Range Image and Video GenerationChain of Thought Empowers Transformers to Solve Inherently Serial ProblemsKAN: Kolmogorov-Arnold NetworksUS lawmakers unveil bill to make it easier to restrict exports of AI modelsOpenAI’s Model Spec outlines some basic rules for AIRobot dogs armed with AI-targeting rifles undergo US Marines Special Ops evaluationOpenAI Releases ‘Deepfake’ Detector to Disinformation ResearchersAudible’s Test of AI-Voiced Audiobooks Tops 40,000 TitlesTikTok will automatically label AI-generated content created on platforms like DALL·E 3Katy Perry's Fan-Made AI Image Is So Real It Fooled the World Into Thinking She Was at the Met GalaSouth Korean woman falls for deepfake Elon Musk, loses $50K in romance scamWhy young Russian women appear so eager to marry Chinese men

Duration:01:41:24

Ask host to enable sharing for playback control

#165 - Sora challenger, Astribot's S1, Med-Gemini, Refusal in LLMs

5/5/2024
Our 165th episode with a summary and discussion of last week's big AI news! Read out our text newsletter and comment on the podcast at https://lastweekin.ai/ Email us your questions and feedback at contact@lastweekin.ai and/or hello@gladstone.ai Timestamps + links: GitHub releases an AI-powered tool aiming for a 'radically new way of building software'China unveils Sora challenger able to produce videos from text similar to OpenAI tool, though much shorterChatGPT’s AI ‘memory’ can remember the preferences of paying customersRabbit R1 review: Avoid this AI gadgetAmazon Q, a generative AI-powered assistant for businesses and developers, is now generally availableYelp’s Assistant AI bot will do all the talking to help users find service providersVideo of super-fast, super-smooth humanoid robot will drop your jawTesla’s 2 million car Autopilot recall is now under federal scrutinyTesla shares soar as Elon Musk returns from China with FSD 'Game Changer'OpenAI inks strategic tie-up with UK’s Financial Times, including content useOpenAI Startup Fund quietly raises $15MHuawei backs HBM memory manufacturing in China to sidestep crippling US sanctions that restrict AI developmentCapabilities of Gemini Models in MedicineLet's Think Dot by Dot: Hidden Computation in Transformer Language ModelsNExT: Teaching Large Language Models to Reason about Code ExecutionSenseNova 5.0: China’s latest AI model surpasses OpenAI’s GPT-4Octopus v4: Graph of language modelsBetter & Faster Large Language Models via Multi-token PredictionRefusal in LLMs is mediated by a single directionRishi Sunak promised to make AI safe. Big Tech’s not playing ball.DOE Announces New Actions to Enhance America’s Global Leadership in Artificial IntelligenceThe Chips Act is rebuilding US semiconductor manufacturing, so far resulting in $327 billion in announced projectsAnalysis-Second global AI safety summit faces tough questions, lower turnoutSam Altman, Jensen Huang, and more join the federal AI safety boardAir Head creators say OpenAI's Sora finicky to work with, needs hundreds of prompts, serious VFX work for under 2 minutes of cohesive story ↺Eight newspaper publishers sue OpenAI over copyright infringement

Duration:01:32:46

Ask host to enable sharing for playback control

#164 - Meta AI, Phi-3, OpenELM, Bollywood Deepfakes

4/30/2024
Our 164th episode with a summary and discussion of last week's big AI news! Read out our text newsletter and comment on the podcast at https://lastweekin.ai/ Email us your questions and feedback at contact@lastweekin.ai and/or hello@gladstone.ai Timestamps + links: Meta, in Its Biggest A.I. Push, Places Smart Assistants Across Its AppsMicrosoft launches Phi-3, its smallest AI model yetThe Ray-Ban Meta Smart Glasses have multimodal AI nowOpenAI winds down AI image generator that blew minds and forged friendships in 2022Baidu claims 200 million users for Ernie chatbot after only 13 monthsThe new Adobe Photoshop gets an in-app image generator, major Generative Fill upgradesIntel & The Pentagon Deepen Ties To Develop World’s Most Advanced ChipsMeta Says It Plans to Spend Billions More on A.I.OpenAI CEO Sam Altman invests in solar power firm Exowatt to fuel AI datacentersGoogle consolidates AI-focused DeepMind, Research teamsMicrosoft and OpenAI bet $100 billion to free themselves from the shackles and overreliance on the world's most profitable semiconductor chip brand for AI chipsApple releases OpenELM: small, open source AI models designed to run on-deviceSnowflake launches Arctic, an open ‘mixture-of-experts’ LLM to take on DBRX, Llama 3The Instruction Hierarchy: Training LLMs to Prioritize Privileged InstructionsGroq’s breakthrough AI chip achieves blistering 800 tokens per second on Meta’s LLaMA 3Microsoft shows off VASA-1, an AI framework that makes human headshots talk, singIntel Builds World’s Largest Neuromorphic System to Enable More Sustainable AIDeepfakes of Bollywood stars spark worries of AI meddling in India electionLLM Agents can Autonomously Exploit One-day VulnerabilitiesThe Necessity of AI Audit Standards BoardsA Mechanism-Based Approach to Mitigating Harms from Persuasive Generative AICOERCING LLMS TO DO AND REVEAL (ALMOST) ANYTHINGChina acquired recently banned Nvidia chips in Super Micro, Dell servers, tenders showDrake threatened with lawsuit over diss track featuring AI Tupac

Duration:01:31:56

Ask host to enable sharing for playback control

#163 - Llama 3, Grok-1.5 Vision, new Atlas robot, RHO-1, Medium ban

4/24/2024
Our 163rd episode with a summary and discussion of last week's big AI news! Note: apology for this one coming out a few days late, got delayed in editing it -Andrey Read out our text newsletter and comment on the podcast at https://lastweekin.ai/ Email us your questions and feedback at contact@lastweekin.ai and/or hello@gladstone.ai Timestamps + links: Meta releases Llama 3, claims it’s among the best open models availableElon Musk’s xAI Unveils Grok-1.5 Vision, Beats OpenAI’s GPT-4VReka releases Reka Core, its multimodal language model to rival GPT-4 and Claude 3 OpusCohere Compass Private Beta: A New Multi-Aspect Embedding ModelAmazon Music’s Maestro lets listeners make AI playlistsSnap plans to add watermarks to images created with its AI-powered toolsBoston Dynamics unveils new Atlas robot for commercial useTSMC’s $65 billion bet still leaves US missing piece of chip puzzleU.S. blacklists Intel's and Nvidia's key partner in China — three other Chinese firms also included in the blacklist for helping the militaryElon Musk says the next-generation Grok 3 model will require 100,000 Nvidia H100 GPUs to trainDr. Andrew Ng appointed to Amazon’s Board of DirectorsCollaborative Robotics Locks Up $100M, Latest Robot Startup To Raise BigOpenEQA: Embodied Question Answering in the Era of Foundation ModelsIntroducing Idefics2: A Powerful 8B Vision-Language Model for the communityRHO-1: Not All Tokens Are What You NeedScaling Laws for Fine-Grained Mixture of ExpertsChinchilla Scaling: A replication attemptChina develops new light-based chiplet that could power artificial general intelligence — where AI is smarter than humansOSWorld: Benchmarking Multimodal Agents for Open-Ended Tasks in Real Computer EnvironmentsU.S. Commerce Secretary Gina Raimondo Announces Expansion of U.S. AI Safety Institute Leadership TeamNSA Publishes Guidance for Strengthening AI System SecurityFoundational Challenges in Assuring Alignment and Safety of Large Language ModelsFormer OpenAI Board Member Calls for Audits of Top AI CompaniesSoA survey reveals a third of translators and quarter of illustrators losing work to AIMedium bans AI-generated content from its paid Partner Program

Duration:01:33:54

Ask host to enable sharing for playback control

#162 - Udio Song AI, TPU v5, Mixtral 8x22, Mixture-of-Depths, Musicians sign open letter

4/14/2024
Our 162nd episode with a summary and discussion of last week's big AI news! Read out our text newsletter and comment on the podcast at https://lastweekin.ai/ Email us your questions and feedback at contact@lastweekin.ai and/or hello@gladstone.ai Timestamps + links: AI-Music Arms Race: Meet Udio, the Other ChatGPT for MusicAnthropic launches external tool use for Claude AI, enabling stock ticker integrations and moreBuilding LLMs for Code RepairEarly Reviews of Humane AI Pin Aren’t ImpressedMicrosoft 365’s Copilot gets a GPT-4 Turbo upgrade and improved image generationAI editing tools are coming to all Google Photos usersGoogle announces the Cloud TPU v5p, its most powerful AI accelerator yetMeta unveils its newest custom AI chip as it races to catch upIntel Unveils New AI Accelerator in Bid to Challenge NvidiaAdobe Is Buying Videos for $3 Per Minute to Build AI ModelOpenAI transcribed over a million hours of YouTube videos to train GPT-4Waymo will launch paid robotaxi service in Los Angeles on WednesdayOpenAI removes Sam Altman's ownership of its Startup FundMistral AI Stuns With Surprise Launch of New Mixtral 8x22B ModelGoogle updates its Gemma AI model family with variants for coding and researchAurora-M: The First Open Source Multilingual Language Model Red-teamed according to the U.S. Executive OrderMixture-of-Depths: Dynamically allocating compute in transformer-based language modelsLeave No Context Behind: Efficient Infinite Context Transformers with Infini-attentionOctopus v2: On-device language model for super agentBigger is not Always Better: Scaling Properties of Latent Diffusion ModelsMany-shot JailbreakingSchiff unveils AI training transparency measureLinwei Ding was a Google software engineer. He was also a prolific thief of trade secrets, say prosecutors.Responsible Reporting for Frontier AI DevelopmentUS govt wants to talk to tech companies about AI electricity demands — eyes nuclear fusion and fissionWashington state judge blocks use of AI-enhanced video as evidence in possible first-of-its-kind rulingTrudeau announces $2.4 billion for AI-related investmentsBillie Eilish, Pearl Jam, Nicki Minaj Among 200 Artists Calling for Responsible AI Music PracticesOpenAI's Sora just made its first music video and it's like a psychedelic trip

Duration:01:45:00

Ask host to enable sharing for playback control

#161 - Claude 3 beats GPT-4, Stability CEO resigns, DBRX, TacticAI, UN resolution on AI

3/31/2024
Our 161st episode with a summary and discussion of last week's big AI news! Check out our sponsor, the SuperDataScience podcast. You can listen to SDS across all major podcasting platforms (e.g., Spotify, Apple Podcasts, Google Podcasts) plus there’s a video version on YouTube. Read out our text newsletter and comment on the podcast at https://lastweekin.ai/ Email us your questions and feedback at contact@lastweekin.ai and/or hello@gladstone.ai Note - one extra story we didn't get to but worth knowing from this week: ‘Totally surreal’: OpenAI shares first short films created with new AI tool Sora Timestamps + links: Google starts testing AI overviews from SGE in main Google search interfaceAdobe’s new GenStudio platform is an AI factory for advertisersClaude-3 Haiku has reached GPT-4 level by our user preferenceMicrosoft Teams is getting smarter Copilot AI featuresSamsung is beating Apple in the race to bring AI to smartphonesElon Musk says all premium subscribers on X will gain access to AI chatbot Grok this weekStability AI CEO resigns to ‘pursue decentralized AI’Amazon spends $2.75 billion on AI startup Anthropic in its largest venture investment yetIntel Gaudi 2 Accelerators Showcase Competitive Performance Per Dollar Against NVIDIA H100 In MLPerf 4.0 GenAI BenchmarksChip Startup Celestial AI Lands Massive $175M Series CAccenture Invests in Sanctuary AI to Bring AI-Powered, Humanoid Robotics to Work Alongside HumansHumanoid robots are joining the Mercedes-Benz workforceIntroducing DBRX: A New State-of-the-Art Open LLMCommon Corpus: A Large Public Domain Dataset for Training LLMsDROID: A Large-Scale In-the-Wild Robot Manipulation DatasetInternLM2 Technical ReportTacticAI: an AI assistant for football tacticsOn the Conversational Persuasiveness of Large Language Models: A Randomized Controlled TrialAutoDev: Automated AI-Driven DevelopmentReverse Training to Nurse the Reversal CurseCobra: Extending Mamba to Multi-Modal Large Language Model for Efficient InferenceGamba: Marry Gaussian Splatting with Mamba for single view 3D reconstructionGeneral Assembly adopts landmark resolution on artificial intelligenceIsrael Deploys Expansive Facial Recognition Program in GazaLINEARITY OF RELATION DECODING IN TRANSFORMER LANGUAGE MODELSUS Weighs Sanctioning Huawei’s Secretive Chinese Chip NetworkNew York City welcomes robotaxis — but only with safety driversThe White House Puts New Guardrails on Government Use of AIBBC Will Stop Using AI For ‘Doctor Who’ Promotion After Receiving Complaints

Duration:01:36:24

Ask host to enable sharing for playback control

#160 - Nvidia's new GPU, Microsoft pays for Inflection AI, Grok-1 open sourced, Jeremie's Action Plan

3/24/2024
Our 160th episode with a summary and discussion of last week's big AI news! Read out our text newsletter and comment on the podcast at https://lastweekin.ai/ Email us your questions and feedback at contact@lastweekin.ai and/or hello@gladstone.ai Timestamps + links: Adobe Substance 3D’s AI features can turn text into backgrounds and texturesOpenAI’s chatbot store is filling up with spamApple’s AI ambitions could include Google or OpenAINvidia reveals Blackwell B200 GPU, the ‘world’s most powerful chip’ for AIMicrosoft to Pay Inflection AI $650 Million After Scooping Up Most of StaffFigure 01: Conversations & Actions in Humanoid Robotics!OpenAI's GPT-4.5 Turbo leaked on search engines and could launch in JuneAbu Dhabi in talks to invest in OpenAI chip ventureNvidia Announces GR00T, a Foundation Model For HumanoidsOpen Release of Grok-1Stability AI brings a new dimension to video with Stable Video 3DColossal-AI Team Introduces Open-Sora: An Open-Source Library for Video GenerationEvolutionary Optimization of Model Merging RecipesDiPaCo: Distributed Path CompositionMM1: Methods, Analysis & Insights from Multimodal LLM Pre-trainingPERL: Parameter Efficient Reinforcement Learning from Human FeedbackVideoAgent: Long-form Video Understanding with Large Language Model as AgentMusicHiFi: Fast High-Fidelity Stereo VocodingThe WMDP Benchmark: Measuring and Reducing Malicious Use With UnlearningExclusive: U.S. Must Move ‘Decisively’ to Avert ‘Extinction-Level’ Threat From AI, Government-Commissioned Report SaysEvaluating Frontier Models for Dangerous CapabilitiesChinese and western scientists identify ‘red lines’ on AI risksGoogle fined $272M by French government over AI use of news contentElvis Act Signed Into Tennessee Law to Protect Musicians From AI DeepfakesAI-Generated ScienceYouTube adds new AI-generated content labeling tool10 of My Most Popular Text-To-Image Series (+Prompts)

Duration:01:39:34

Ask host to enable sharing for playback control

#159 - Inflection-2.5, Devin, OpenAI board update, SIMA, EU AI Act passed

3/18/2024
Our 159th episode with a summary and discussion of last week's big AI news! Check out our sponsor, the SuperDataScience podcast. You can listen to SDS across all major podcasting platforms (e.g., Spotify, Apple Podcasts, Google Podcasts) plus there’s a video version on YouTube. Read out our text newsletter and comment on the podcast at https://lastweekin.ai/ Email us your questions and feedback at contact@lastweekin.ai and/or hello@gladstone.ai Timestamps + links: Inflection-2.5: meet the world's best personal AIIntroducing Devin, the first AI software engineerDoorDash’s new AI-powered ‘SafeChat+’ tool automatically detects verbal abuseAnthropic releases Claude 3 Haiku, an AI model built for speed and affordabilityPika Labs just added sound effects to its generative AI videos — here’s how it soundsSalesforce announces new AI tools for doctorsSam Altman Rejoins OpenAI Board Along With Three New DirectorsCohere releases powerful ‘Command-R’ language model for enterprise useBuilding Meta’s GenAI InfrastructureBaidu Launches China's First 24/7 Robotaxi ServiceCroissant: a metadata format for ML-ready datasetsSaulLM-7B: A pioneering Large Language Model for LawKai-Fu Lee’s AI Company “01.AI” Announces the Open Source of the Yi-9B ModelA generalist AI agent for 3D virtual environmentsStealing Part of a Production Language ModelData Interpreter: An LLM Agent For Data ScienceShortGPT: Layers in Large Language Models are More Redundant Than You ExpectPixArt-Σ: Weak-to-Strong Training of Diffusion Transformer for 4K Text-to-Image GenerationWorld’s first major act to regulate AI passed by European lawmakersUS spearheads first UN resolution on artificial intelligence — aimed at ensuring equal accessGoogle restricts election-related queries for its Gemini chatbotResearchers tested leading AI models for copyright infringement using popular books, and GPT-4 performed worstNvidia Says NeMo AI Platform Complies With Copyright After Authors’ ComplaintFive of this year’s Pulitzer finalists are AI-poweredI made by Superman action figure talk with Pika Labs’ new AI lip sync tool — watch this

Duration:01:00:03

Ask host to enable sharing for playback control

#158 - Claude 3, Elon Musk sues OpenAI, StarCoder 2, AI-Generated Spam

3/10/2024
Our 158th episode with a summary and discussion of last week's big AI news! Read out our text newsletter and comment on the podcast at https://lastweekin.ai/ Email us your questions and feedback at contact@lastweekin.ai and/or hello@gladstone.ai Timestamps + links: Introducing the next generation of ClaudeCompetition in AI video generation heats up as Deepmind alums unveil HaiperMeta AI creates ahistorical images, like Google GeminiIdeogram Is A New AI Image Generator That Obliterates the Competition, Outperforming MidJourney and Dall-E 3Wix’s new AI chatbot builds websites in seconds based on promptsI used generative AI to turn my story into a comic—and you can tooElon Musk sues OpenAI and CEO Sam Altman for putting profits above humanityOpenAI Fires Back at Musk Allegations With Trove of EmailsInside the Crisis at GoogleIt’s official: Waymo robotaxis are now free to use freeways and leave San FranciscoNvidia's next-gen AI GPUs could draw an astounding 1000 Watts each, a 40 percent increase — Dell spills the beans on B100 and B200 in its earnings callAI chip startup Groq forms new business unit, acquires Definitive IntelligenceStarCoder 2 and The Stack v2: The Next GenerationIntroducing TripoSR: Fast 3D Object Generation from Single ImagesH2O AI releases Danube, a super-tiny LLM for mobile applicationsAtP*: An efficient and scalable method for localizing LLM behaviour to componentsStable Diffusion 3: Research PaperApproaching Human-Level Forecasting with Language ModelsHere Come the AI WormsHigh-speed humanoid feels like a step change in roboticsFunctional Benchmarks for Robust Evaluation of Reasoning Performance, and the Reasoning GapIndia reverses AI stance, requires government approval for model launchesWhen Your AIs Deceive You: Challenges with Partial Observability of Human Evaluators in Reward LearningOpenAI signs open letter to build AI responsibly just days after Elon Musk sued the company for putting profit ahead of peopleAI-generated articles prompt Wikipedia to downgrade CNET’s reliability ratingMalicious AI models on Hugging Face backdoor users’ machinesChina offers AI computing ‘vouchers’ to its underpowered start-upsTrump supporters target black voters with faked AI imagesAI-Generated Kara Swisher Biographies Flood AmazonInside the World of AI TikTok SpammersTwitter is becoming a 'ghost town' of bots as AI-generated spam content floods the internetMan tries to steal driverless car in L.A., doesn’t get far: police

Duration:01:48:45

Ask host to enable sharing for playback control

#157 - Gemini controversy, new Mistral models, Deepmind's Genie & Griffinn, AI Warfare is here

3/3/2024
Our 157th episode with a summary and discussion of last week's big AI news! Check out our sponsor, the SuperDataScience podcast. You can listen to SDS across all major podcasting platforms (e.g., Spotify, Apple Podcasts, Google Podcasts) plus there’s a video version on YouTube. Bonus plug: also check out this new book by Stanford AI expert, bestselling author, and Last Week in AI supporter Jerry Kaplan! Generative Artificial Intelligence: What Everyone Needs to Know Read out our text newsletter and comment on the podcast at https://lastweekin.ai/ Email us your questions and feedback at contact@lastweekin.ai and/or hello@gladstone.ai Timestamps + links: Google apologizes for ‘missing the mark’ after Gemini generated racially diverse NazisStability announces Stable Diffusion 3, a next-gen AI image generatorMistral AI releases new model to rival GPT-4 and its own chat assistantWindows just got its own Magic Eraser to AI-modify your photosAdobe previews new cutting-edge generative AI tools for crafting and editing custom audioAI video wars heat up as Pika adds Lip Sync powered by ElevenLabsMicrosoft Strikes Deal with France’s Mistral AIFigure Raises $675M at $2.6B Valuation and Signs Collaboration Agreement with OpenAINvidia posts record revenue up 265% on booming AI businessMediaTek’s latest chipsets are now ‘optimized’ for Gemini NanoTumblr’s owner is striking deals with OpenAI and Midjourney for training data, says reportMistral AI models coming soon to Amazon BedrockGenerative AI Startup Mistral Releases Free ‘Open-Source’ 7.3B Parameter LLMGoogle Delves Deeper Into Open Source with Launch of Gemma AI ModelMicrosoft releases its internal generative AI red teaming tool to the publicIntroducing Phind-70B – closing the code quality gap with GPT-4 Turbo while running 4x fasterGenie: Generative Interactive EnvironmentsGriffin: Mixing Gated Linear Recurrences with Local Attention for Efficient Language ModelsQuantum Circuit Optimization with AlphaTensorBack to Basics: Revisiting REINFORCE Style Optimization for Learning from Human Feedback in LLMsRepetition Improves Language Model EmbeddingsAI Warfare Is Already HereMan admits to paying magician $150 to create anti-Biden robocallGoogle DeepMind forms a new org focused on AI safetyFacebook whistleblower, AI godfather join hundreds calling for deepfake regulationUS regulators investigate whether OpenAI investors were misled, say reportsUsers Say Microsoft's AI Has Alternate Personality as Godlike AGI That Demands to Be WorshippedThe Intercept, Raw Story, and AlterNet sue OpenAI and MicrosoftA viral photo of a guy smoking in McDonald's is completely fake — and of course made by AIImpossible AI Food

Duration:01:44:57

Ask host to enable sharing for playback control

#156 - OpenAI's Sora, Gemini 1.5, BioMistral, V-JEPA, AI Task Force, Fun!

2/25/2024
Our 156th episode with a summary and discussion of last week's big AI news! Read out our text newsletter and comment on the podcast at https://lastweekin.ai/ Email us your questions and feedback at contact@lastweekin.ai and/or hello@gladstone.ai Timestamps + links: OpenAI introduces Sora, its text-to-video AI modelGemini 1.5 is Google’s next-gen AI model — and it’s already almost readyGroq AI model goes viral and rivals ChatGPT, challenges Elon Musk’s GrokIntroducing IP Adapters: Create Consistent Game Assets in SecondsReport: OpenAI working on web search productAdobe Acrobat adds generative AI to ‘easily chat with documents’Sam Altman owns OpenAI's venture capital fundReddit Signs AI Content Licensing Deal Ahead of IPONvidia reveals its Eos supercomputer for AI processing sporting 4,608 H100 GPUsGoogle quietly launches internal AI model named 'Goose' to help employees write code faster, leaked documents showChinese start-up Moonshot AI raises US$1 billion in funding round led by Alibaba and VC HongShan amid strong interest for OpenAI-type firmsAI Computing Firm Lambda Raises $320 Million in Fresh FundingExclusive: Ex-Salesforce Co-CEO Bret Taylor and longtime Googler Clay Bavor raised $110 million to bring AI ‘agents’ to businessBioMistral: A Collection of Open-Source Pretrained Large Language Models for Medical DomainsNomic AI Releases the First Fully Open-Source Long Context Text Embedding Model that Surpasses OpenAI Ada-002 Performance on Various BenchmarksMeta unveils V-JEPA AI model that improves training by learning from videoChain-of-Thought Reasoning Without PromptingOS-Copilot: Towards Generalist Computer Agents with Self-ImprovementWorld Model on Million-Length Video And Language With RingAttentionAmazon AGI Team Say Their AI Is Showing "Emergent Abilities"Hackers for China, Russia and Others Used OpenAI Systems, Report SaysHouse leaders launch bipartisan artificial intelligence task forceYour fingerprints can be recreated from the sounds made when you swipe on a touchscreen — Chinese and US researchers show new side channel can reproduce fingerprints to enable attacksStates are introducing 50 AI-related bills per weekAir Canada found liable for chatbot's bad advice on plane ticketsThe FTC warned about ‘quiet’ TOS changes for AI training. Here’s why it might not be enoughSarah Silverman’s lawsuit against OpenAI partially dismissedA Visual Guide to Mamba and State Space ModelsScientific Journal Publishes AI-Generated Rat with Gigantic Penis In Worrying IncidentHelen Mirren Rips Up AI-Generated Speech at American Cinematheque AwardsMicrosoft's game-changing Super Bowl ad

Duration:01:45:06

Ask host to enable sharing for playback control

#155 - ChatGPT memory, Altman seeks trillions, Califonia AI regulation, art gen lawsuit

2/16/2024
Our 155th episode with a summary and discussion of last week's big AI news! Correction: Andrey said CLIP came out with DALL-E 2; it came out alongside the first DALL-E. Check out our sponsor, the SuperDataScience podcast. You can listen to SDS across all major podcasting platforms (e.g., Spotify, Apple Podcasts, Google Podcasts) plus there’s a video version on YouTube. Read out our text newsletter and comment on the podcast at https://lastweekin.ai/ Email us your questions and feedback at contact@lastweekin.ai and/or hello@gladstone.ai Timestamps + links: ChatGPT is getting ‘memory’ to remember who you are and what you likeReka Flash: An Efficient and Capable Multimodal Language ModelSay What? Chat With RTX Brings Custom Chatbot to NVIDIA RTX AI PCsSam Altman Seeks Trillions of Dollars to Reshape Business of Chips and AIReport: China’s SMIC to begin production of 5nm chips for HuaweiA crowd destroyed a driverless Waymo car in San FranciscoOpenAI reportedly developing two AI agents to automate entire work processesCruise names first chief safety officer following crash and controversyCohere for AI launches open source LLM for 101 languagesBUD-E: ENHANCING AI VOICE ASSISTANTS’ CONVERSATIONAL QUALITY, NATURALNESS AND EMPATHYEVA-CLIP-18B: Scaling CLIP to 18 Billion ParametersIntroducing Stable CascadeSELF-DISCOVER: Large Language Models Self-Compose Reasoning StructuresBlackMamba: Mixture of Experts for State-Space ModelsAn Interactive Agent Foundation ModelGrandmaster-Level Chess Without SearchMore Agents Is All You NeedMusicMagus: Zero-Shot Text-to-Music Editing via Diffusion ModelsDebating with More Persuasive LLMs Leads to More Truthful AnswersIn Big Tech’s backyard, California lawmaker unveils landmark AI billAI Deployed Nukes 'to Have Peace in the World' in Tense War SimulationProtesters gather outside OpenAI office, opposing military AI and AGIAI safeguards can easily be broken, UK Safety Institute findsStability, Midjourney, Runway hit back in AI art lawsuitAI Companies Take Hit as Judge Says Artists Have “Public Interest” In Pursuing Lawsuits

Duration:01:45:41

Ask host to enable sharing for playback control

#154 - Google Gemini, Waymo Collision, Smaug-72B, EU AI Act final text, image watermarks

2/11/2024
Our 154th episode with a summary and discussion of last week's big AI news! Read out our text newsletter and comment on the podcast at https://lastweekin.ai/ Email us your questions and feedback at contact@lastweekin.ai and/or hello@gladstone.ai Correction: Andrey mentioned "State space machines", he meant "State space models" Timestamps + links: Google Releases Gemini, an A.I.-Driven Chatbot and Voice AssistantCopilot gets a big redesign and a new way to edit your AI-generated imagesArc Search's AI responses launched as an unfettered experience with no guardrailsBrilliant Labs’s Frame glasses serve as multimodal AI assistantStability AI launches SVD 1.1, a diffusion model for more consistent AI videosOpenAI launches ChatGPT app for Apple Vision ProA Waymo robotaxi hit a cyclist in San Francisco – here’s what happenedCanon plans to disrupt chipmaking with low-cost “stamp” machineUS industry group calls for multilateral chip export controls to address disadvantage over Korea, other alliesU.S. blocks shipment of 24 Nvidia AI GPUs to China over concerns about self-driving truck companyNvidia reportedly selects Intel Foundry Services for GPU packaging production — could produce over 300,000 H100 GPUs per monthAllen Institute for AI launches open and transparent OLMo large language modelMeet ‘Smaug-72B’: The new king of open-source AIIntroducing Qwen1.5Hugging Face launches open source AI assistant maker to rival OpenAI’s custom GPTsApple releases ‘MGIE’, a revolutionary AI model for instruction-based image editingLearning Universal PredictorsCan Mamba Learn How to Learn? A Comparative Study on In-Context Learning TasksMusicRL: Aligning Music Generation to Human PreferencesFP6-LLM: Efficiently Serving Large Language Models Through FP6-CentricAlgorithm-System Co-DesignAgentBoard: An Analytical Evaluation Board of Multi-turn LLM AgentsSpecialized Language Models with Cheap Inference from Limited Domain DataEU’s AI Act passes last big hurdle on the way to adoptionBuilding an early warning system for LLM-aided biological threat creationFCC votes to ban scam robocalls that use AI-generated voicesBiden administration names a director of the new AI Safety InstituteOpenAI's GPT-4 finally meets its match: Scots Gaelic smashes safety guardrailsAI poisoning tool Nightshade received 250,000 downloads in 5 days: ‘beyond anything we imagined’Labeling AI-Generated Images on Facebook, Instagram and ThreadsOpenAI is adding new watermarks to DALL-E 3Following lawsuit, rep admits “AI” George Carlin was human-written

Duration:01:37:16

Ask host to enable sharing for playback control

#153 - Taylor Swift Deepfakes, ChatGPT features, Meta-Prompting, two new US bills

2/4/2024
Our 153rd episode with a summary and discussion of last week's big AI news! Check out our sponsor, the SuperDataScience podcast. You can listen to SDS across all major podcasting platforms (e.g., Spotify, Apple Podcasts, Google Podcasts) plus there’s a video version on YouTube. Read out our text newsletter and comment on the podcast at https://lastweekin.ai/ Email us your questions and feedback at contact@lastweekin.ai and/or hello@gladstone.ai Timestamps + links: Trolls have flooded X with graphic Taylor Swift AI fakesYouTube Deletes 1,000 Videos of Celebrity AI Scam AdsIceland Has Its Own AI George Carlin Moment, Considers Law Against Deepfaking the DeadGuns N’ Roses share AI-generated video for ‘The General’Microsoft Makes Swift Changes to AI ToolOpenAI drops prices and fixes ‘lazy’ GPT-4 that refused to workChatGPT's new "@GPT" feature paves the way for OpenAI's vision of a universal assistantChatGPT finally has competition — Google Bard with Gemini just matched it with a huge upgradeBrowser alternatives Brave, Arc, add new AI integrationsBaidu's Ernie AI chatbot to power Samsung's new Galaxy S24 smartphonesAI companies lose $190 billion in market cap after Alphabet and Microsoft reportCruise wasn’t hiding the pedestrian-dragging video from regulators — it just had bad internetHugging Face teams up with Google to accelerate open AI developmentElon Musk's xAI eyes $6 billion funding to challenge ChatGPT maker OpenAIAI Chip Startup Rebellions Snags Funding to Challenge NvidiaNew open source AI coding tool surpasses its inspiration: Google DeepMind’s AlphaCodeMeta’s free Code Llama AI programming tool closes the gap with GPT-4DeepSeek-Coder: When the Large Language Model Meets Programming -- The Rise of Code IntelligenceMeta-Prompting: Enhancing Language Models with Task-Agnostic ScaffoldingEAGLE: Speculative Sampling Requires Rethinking Feature UncertaintyAre Vision Transformers More Data Hungry Than Newborn Visual Systems?Circuit Component Reuse Across Tasks in Transformer Language ModelsA ‘Shocking’ Amount of the Web Is Already AI-Translated Trash, Scientists DetermineTaylor Swift AI images prompt US bill to tackle nonconsensual, sexual deepfakesOpenAI and Google will be required to notify the government about AI modelsThe campaign to take down the Biden AI executive orderRep. Jeff Jackson Introduces Bipartisan CLOUD AI Act to Stop China from Remotely Using American Technology to Build AI ToolsAI Startup ElevenLabs Bans Account Blamed for Biden Audio DeepfakeHow the West Can Match Russia in Drone Innovation

Duration:01:46:07

Ask host to enable sharing for playback control

#152 - live translation on phones, Meta aims at AGI, AlphaGeometry, political deepfakes

1/28/2024
Our 152nd episode with a summary and discussion of last week's big AI news! Read out our text newsletter and comment on the podcast at https://lastweekin.ai/ Email us your questions and feedback at contact@lastweekin.ai and/or jeremie@gladstone.ai Timestamps + links: Samsung’s latest Galaxy phones offer live translation over phone calls, textsThe Rabbit R1 will receive live info from Perplexity’s AI ‘answer engine’Google is using AI to organize and customize your Chrome browserAdobe’s new AI-powered Premiere Pro features eradicate boring audio editing tasksWaymo looks to launch full fleet of robotaxis in LATesla finally releases FSD v12, its last hope for self-drivingOpenAI CEO Sam Altman is still chasing billions to build AI chipsMark Zuckerberg’s new goal is creating artificial general intelligenceAI voice startup ElevenLabs lands $80M round, launches marketplace of cloned voicesAnthropic’s Margins Raise Questions on AI Startups’ Long-Term ProfitabilityCohere in talks to raise as much as $1bn as AI arms race heats upDOJ and FTC push to investigate Microsoft's OpenAI partnership- PoliticoFigure announces commercial agreement with BMW Manufacturing to bring general purpose robots into automotive productionStability AI unveils smaller, more efficient 1.6B language model as part of ongoing innovationAlphaGeometry: An Olympiad-level AI system for geometryLumiere: A Space-Time Diffusion Model for Video GenerationChatQA: Building GPT-4 Level Conversational QA ModelsVision Mamba: Efficient Visual Representation Learning with Bidirectional State Space ModelVMamba: Visual State Space ModelDepth Anything: Unleashing the Power of Large-Scale Unlabeled DataOpenAI suspends developer behind Dean Phillips botFake Joe Biden robocall tells New Hampshire Democrats not to vote TuesdaySharing Fake Nude Images Could Become a Federal Crime Under Proposed LawSan Francisco takes legal action over ‘unsafe,’ ‘disruptive’ self-driving carsSlew of deepfake video adverts of Sunak on Facebook raises alarm over AI risk to electionAI is the buzz, the big opportunity and the risk to watch among the Davos glitteratiTest Yourself: Which Faces Were Made by A.I.?AI models that don’t violate copyright are getting a new certification label

Duration:01:27:18

Ask host to enable sharing for playback control

#151 - Copilot Pro, LLama.cpp, conversational diagnostic AI, secret AI diplomacy

1/21/2024
Our 151st episode with a summary and discussion of last week's big AI news! Check out our sponsor, the SuperDataScience podcast. You can listen to SDS across all major podcasting platforms (e.g., Spotify, Apple Podcasts, Google Podcasts) plus there’s a video version on YouTube. Read out our text newsletter and comment on the podcast at https://lastweekin.ai/ Email us your questions and feedback at contact@lastweekin.ai Timestamps + links: Microsoft’s new Copilot Pro brings AI-powered Office features to the rest of usMicrosoft Copilot is now using the previously-paywalled GPT-4 Turbo, saving you $20 a monthAmazon launches generative AI tool to answer shoppers’ questionsDeepMind spin-off aims to halve drug discovery times following Big Pharma dealsChina planning 1,600-core chips that use an entire wafer — similar to American company Cerebras 'wafer-scale' designsChatGPT will have video functionality and more accuracy in future versions — Sam Altman says GPT-5 will be a big improvementHuawei Teardown Shows 5nm Chip Made in Taiwan, Not ChinaOpenAI’s news publisher deals reportedly top out at $5 million a yearInside AI Unicorn Anthropic’s Unusual $750 Million FundraiseMicrosoft executive Dee Templeton joins OpenAI board - Bloomberg NewsOpenAI-backed 1X raises another $100M for the race to humanoid robotsMeet LLama.cpp: An Open-Source Machine Learning Library to Run the LLaMA Model Using 4-bit Integer Quantization on a MacBookLLAMA PRO: Progressive LLaMA with Block ExpansionOne of the world’s largest AI training datasets is about to get bigger and ‘substantially better’Towards Conversational Diagnostic AISleeper Agents: Training Deceptive LLMs that Persist Through Safety TrainingLLM Augmented LLMs: Expanding Capabilities through CompositionSelf-Play Fine-Tuning Converts Weak Language Models to Strong Language ModelsClinical predictive models created by AI are accurate but study-specific, researchers findSecure, Governable ChipsUS companies and Chinese experts engaged in secret diplomacy on AI safetyChina Cyberspace Security Association releases first batch of Chinese basic corpusOpenAI Bans Use of AI Tools for Campaigning, Voter SuppressionOpenAI Quietly Deletes Ban on Using ChatGPT for “Military and Warfare”Musicians Set to Begin Contract Negotiations With Studios On AI, Streaming PrioritiesScammy AI-Generated Book Rewrites Are Flooding AmazonDeepfaked Celebrity Ads Promoting Medicare Scams Run Rampant on YouTube

Duration:01:36:10