Mixture of Experts

Are browsers the right entry point for AI tools? In episode 71 of Mixture of Experts, guest host Bryan Casey is joined by Gabe Goodhart, Kaoutar El Maghraoui and Mihai Criveti to talk about the verdict in the Google antitrust case and what it means for agentic AI. Next, as Anthropic raised  $13 billion in a recent funding round, bringing its valuation to $183 billion, we discuss investment in AI startups. Finally, the discourse on GPT-5 and AI model innovation created “AI winter.” What does this mean for the future of AI innovation? All that and more on today’s Mixture of Experts.  00:00 – Intro 1:38 – Safer GPT, IBM and AMD, Amazon Lens Live and AI and Starbucks 2:59 – Google antitrust 27:08 – Anthropic and AI valuation 41:42 – AI winter  The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity. Subscribe for AI updates → https://www.ibm.com/account/reg/us-en/signup?formid=news-urx-52120 Learn more about artificial intelligence → https://www.ibm.com/think/artificial-intelligence Visit Mixture of Experts podcast page to get more AI content → https://www.ibm.com/think/podcasts/mixture-of-experts 

Would you trust a 100-page prompt to do your taxes? In episode 70 of Mixture of Experts, host Tim Hwang is joined by Aaron Baughman, Chris Hay and Lauren McHugh to talk about KPMG’s 100-page prompt they used to build their agentic TaxBot. Next, we debrief on OpenAI’s teasing of selling infrastructure in the future. The image model generation goes bananas, with the latest AI Image Model from Gemini: nano-banana. Finally, Aaron Baughman demonstrated three new features for the US Open website, powered by IBM watsonx-built generative AI models: Match Chat, Key Points, and Live Likelihood to Win. All that and more on today’s 70th episode of Mixture of Experts.00:00 – Intro 3:05 – KPMG’s monster prompt  16:37 – OpenAI’s infra for sale? 25:10 – Gemini's nano-banana 35:11 – US Open experimentations   The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity. Subscribe for AI updates → https://www.ibm.com/account/reg/us-en/signup?formid=news-urx-52120 Learn more about artificial intelligence → https://www.ibm.com/think/artificial-intelligence Visit Mixture of Experts podcast page to get more AI content → https://www.ibm.com/think/podcasts/mixture-of-experts US Open powered by IBM Watsonx  https://www.ibm.com/sports/usopen 

Is enterprise AI in danger? In episode 69 of Mixture of Experts, host Tim Hwang is joined by Marina Danilevsky, Nathalie Baracaldo and Sandi Besen to debrief MIT’s report on gen AI pilots. Next, GPT-5 has a hidden system prompt? Then, we revisit the conversation about chain of thought (CoT) reasoning with our researchers. Are large reasoning models not thinking straight? Finally, Anthropic announced Claude will close down "distressing” conversations and we debate AI welfare. All that and more on today’s episode of Mixture of Experts. 00:00 – Intro 1:13 – US Open, Meta restructuring Superintelligence lab and Robot Olympics 3:11 – Gen AI pilots fail 11:09 – GPT-5's hidden prompt revealed 22:47 – Reasoning model flaws 33:55 – Claude closing chats  The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity. Subscribe to the Think newsletter → https://www.ibm.com/account/reg/us-en/signup?formid=news-urx-52120 Learn more about artificial intelligence → https://www.ibm.com/think/artificial-intelligence Visit Mixture of Experts podcast page to get more AI content → https://www.ibm.com/think/podcasts/mixture-of-experts 

Would you sell Chrome for USD 34.5 billion dollars? In episode 68 of Mixture of Experts, host Tim Hwang is joined by Abraham Daniels, Sophie Kuijt and Shobhit Varshney for another packed week in AI. First, AI startup Perplexity puts out a bid for Google Chrome at over double their valuation. Why? Next, xAI released Grok Imagine and claims it will be the next Vine. Our experts analyze the future of AI video generation. Finally, one week after the GPT-5 release and skeptics are saying it did not live up to the hype. Is AI development plateauing? All that and more on Mixture of Experts! 00:00 – Intro 01:17 – MoE News: NVIDIA H20s, Apple AI devices, AI people pleasers and Google DeepMind’s bioacoustics model 02:40 – Perplexity’s USD 34.5B bid 12:10 – Grok Imagine 24:23 – GPT-5 check-in The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity. Subscribe to the Think Newsletter → https://www.ibm.com/account/reg/us-en/signup?formid=news-urx-52120 Learn more about artificial intelligence → https://www.ibm.com/think/artificial-intelligence Visit Mixture of Experts podcast page to get more AI content → https://www.ibm.com/think/podcasts/mixture-of-experts 

Is GPT-5 better at code than Claude Opus 4.1? In this bonus episode of Mixture of Experts, guest host Bryan Casey is joined by Mihai Criveti and Chris Hay to analyze OpenAI’s new release. Our experts dove into the new models and compared them with their favorites. In today’s special episode, we dive into the news, impacts on the industry and even share a demo. Can GPT-5 outperform Claude Opus 4.1? All that and more on today’s episode of Mixture of Experts.  00:00 – Intro 1:26 -- GPT-5 is here 13:43 -- Are we closer to AGI? 20:50 -- Demo:  Opus 4.1 vs. GPT-5  The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity.  Subscribe for AI updates → https://www.ibm.com/account/reg/us-en/signup?formid=news-urx-52120 Learn more about artificial intelligence → https://www.ibm.com/think/artificial-intelligence Visit Mixture of Experts podcast page to get more AI content → https://www.ibm.com/think/podcasts/mixture-of-experts 

OpenAI goes open-weight?  In episode 67 of Mixture of Experts, host Tim Hwang is joined by Kaoutar El Maghraoui, Chris Hay and Bruno Aziza to debrief OpenAI’s release of gpt-oss, their new open-source models. Next, the model releases continue as Google DeepMind dropped Genie 3. Our experts analyze the various AI model performance. Then, there’s been some drama surrounding the pricing of Claude Code. We’ll discuss where Anthropic landed and what it means. Finally, Mark Zuckerberg shared a new essay on Personal Superintelligence. Our experts take us through what superintelligence looks like across the major AI players. All that and more on today’s episode of Mixture of Experts. 00:00 -- Intro 01:20 -- Gpt-oss 12:26 -- Genie 3 22:12 -- Claude pricing 33:04 -- Zuckerberg on Personal Superintelligence   The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity. Subscribe for AI updates → https://www.ibm.com/account/reg/us-en/signup?formid=news-urx-52120 Learn more about artificial intelligence → https://www.ibm.com/think/artificial-intelligence Visit Mixture of Experts podcast page to get more AI content → https://www.ibm.com/think/podcasts/mixture-of-experts 

Is ChatGPT making you dumb? In episode 66 of Mixture of Experts, host Tim Hwang is joined by Kaoutar El Maghraoui, Kush Varshney and Volkmar Uhlig. First, ChatGPT released a new study mode. The intention is to support education, but what is the reality? Next, AI agents are changing design interfaces; is agentic experience (AX) the new UX? Then, a new paper released by Nature about generative neural networks contextualizing ancient texts. How is AI supporting historical research? Finally, special guest, Suja Viswesan, joins us to debrief the 2025 Cost of a Data Breach Report. What do we need to know about AI-driven cybersecurity attacks? Tune in to Mixture of Experts to find out! 00:00 – Intro 01:09 – ChatGPT study mode 13:52 – Agentic experience 12:08 – Decoding ancient texts with AI 39:55 – Cost of a Data Breach Report 2025 The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity. Read the 2025 Cost of a Data Breach Report → https://www.ibm.com/reports/data-breach  Subscribe for AI updates → https://www.ibm.com/account/reg/us-en/signup?formid=news-urx-52120 Learn more about artificial intelligence → https://www.ibm.com/think/artificial-intelligence Visit Mixture of Experts podcast page to get more AI content → https://www.ibm.com/think/podcasts/mixture-of-experts 

What does the White House have to say about AI? In episode 65 of Mixture of Experts, host Tim Hwang is joined by Kate Soule, Gabe Goodhart  and first-time guest, Mihai Criveti. First, Google DeepMind shared that Gemini Deep Think won Gold at IMO. Next, who is using ChatGPT agents? We get our experts’ thoughts. Then, Mihai takes us through MCP Gateway and what this means for next-gen AI systems. Finally, special guest, Ryan Hagemann, joins us to analyze the White House’s new AI Action Plan, released this week. What does this mean for AI policy? Tune in to Mixture of Experts to find out!  00:00 – Intro 01:16 - DeepMind at IMO 16:27 - ChatGPT agents 25:43 - MCP Gateway 35:45 - AI Action Plan The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity. 

Is Kimi K2 actually better than Claude? In episode 64 of Mixture of Experts, host Tim Hwang is joined by Abraham Daniels, Chris Hay and Kaoutar El Maghraoui. First, Moonshot AI released Kimi K2, their trillion-parameter MoE model, and our experts analyze the benchmarks and what this really means. Then, we reflect on DeepSeek-R1 6 months later; did it live up to the hype? Next, Google is investing $25 billion in AI infrastructure, and it’s not just AI chips. How does this compare to their competitors? Finally, Anthropic’s Claude for Enterprise announced an expansion with Lawrence Livermore National Laboratory —what AI safety concerns might this raise?  Tune in to today’s episode of Mixture of Experts to find out. 00:00 – Intro 01:18 – Kimi K2 12:07 – DeepSeek-R1 vibe check 28:49 – Google's data center investments 41:20 – Claude powers LLNL research  The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity. Resources:Read more on how DeepSeek has changed the landscape of AI, six months after the watershed release of R1 → http://ibm.com/think/news/deepseek-global-ai-local  Subscribe for AI updates → https://www.ibm.com/account/reg/us-en/signup?formid=news-urx-52120 Visit Mixture of Experts podcast page to get more AI content → https://www.ibm.com/think/podcasts/mixture-of-experts 

How will AI agents change search? In episode 63 of Mixture of Experts, host Tim Hwang is joined by Aaron Baughman, Chris Hay and Kate Soule. First, Perplexity released their Comet browser, and there are rumors that OpenAI is next. Who will win the AI browser war? Then, we discuss frontier model transparency amid Anthropic’s call for a “targeted transparency framework”. Does this only support large model developers? Later, Cloudflare blocks AI scrapers. Is this a good thing? Finally, we discuss how AI is showing up at Wimbledon this year and enhancing the fan experience via Match Chat.  Tune in to today’s episode of Mixture of Experts. 00:00 – Intro 1:06 -- AI browser wars 13:10 -- Anthropic on model transparency 28:57 -- Cloudflare scrapers 41:32 -- Wimbledon Match Chat The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity. 

Will AI agents run all businesses? In episode 62 of Mixture of Experts, host Tim Hwang is joined by Gabe Goodhart, Kush Varshney and Marina Danilevsky to debrief Anthropic’s Project Vend. Next, do we still need massive data centers? We analyze DiLoCoX and discuss the possibility of distributed model training. Then, the New York Times released an article discussing how computer science education has changed in the era of AI; should people still study computer science? Finally, is the paper review process broken? And is it AI’s fault? All that and more on today’s episode of Mixture of Experts. 00:00 – Intro 01:07 -- Anthropic’s Project Vend 13:38 -- DiLoCoX 25:56 -- Computer science education 40:57 -- AI prompts in papers The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity. 

In episode 61 of Mixture of Experts, host Tim Hwang is joined by Kaoutar El Maghraoui, Gabe Goodhart and, joining us for the first time, Ann Funai. First up, a new paper from MIT: “Your brain on ChatGPT”. Are we using AI and LLMs to augment our intelligence, or are we becoming optimally lazy? Next, our experts explore the surprising evolution of autonomous vehicles: they are driving more aggressively, and the results might actually be... safer? Finally, a conversation about AI-generated ads, AI-video generation and the risks that come with them.The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity.

Could AI take your job? In episode 60 of Mixture of Experts, host Tim Hwang is joined by Phaedra Boinodiris, Chris Hay and Volkmar Uhlig. First, the impact of AI on the job market is all the rage online. Between the Godfather of AI revealing which jobs he feels are safe, and Jensen Huang responding to Dario Amodei’s thoughts, our experts analyze the chatter. Next, Scale AI is facing some fallout. What can we learn about data security? Then, an article from the New York Times details how chatbots can take users down “conspiratorial rabbit holes,” Who is benefitting from these conversations? Finally, how is AI affecting the startup ecosystem?Tune in to Mixture of Experts to find out!  00:01 – Intro 01:17 -- AI and jobs 12:28 -- Scale AI fallout 22:00 -- Chatbot conspiracies 35:20 -- AI startup ecosystem  The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity. 

Did Apple’s WWDC 2025 live up to expectations? In episode 59 of Mixture of Experts, host Tim Hwang is joined by Chris Hay, Kaoutar El Magrahoui and Shobhit Varshney. Today, the experts analyze all things Apple—from Apple Research’s recent paperThe Illusion of Thinking to Apple Intelligence. Next, OpenAI released o3-pro: we continue the analysis on AI reasoning. Then, Meta purchased Scale AI for a whopping $15 billion. Why? Finally, an exciting new announcement on fault-tolerant quantum computing: IBM Quantum Starling will arrive by 2029. What does this mean and why should we care? All that and more on this week’s Mixture of Experts. 00:01 – Intro 01:58 -- Apple's WWDC  16:47 -- OpenAI o3-pro 30:43 -- Meta & Scale AI's "superintelligence" lab 37:56 -- Fault-tolerant quantum computing  The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity.  Resources:How IBM will build the world's first large-scale, fault-tolerant quantum computer: https://www.ibm.com/quantum/blog/large-scale-ftqcVisit the Mixture of Experts podcast page to learn more: https://www.ibm.com/think/podcasts/mixture-of-expertsSubscribe for AI updates: https://www.ibm.com/account/reg/us-en/signup?formid=news-urx-52120 Learn more about artificial intelligence: https://www.ibm.com/think/artificial-intelligence 

Is open source winning the AI race? In episode 58 of Mixture of Experts, host Tim Hwang is joined by Anthony Annunziata, Ash Minhas and Sarah Amos live from New York Tech Week. First, we dive into the various themes coming out of NY Tech Week, specifically practical uses of AI. Next, we analyze a couple of different reports about the impact of open source on AI. Finally, Claude 4 has some really weird behaviors. What does this teach us about AI safety and model development? All that and more on this week’s Mixture of Experts. 00:01 -- Intro 01:09 -- New York Tech Week 2025 11:36 -- Open source AI reports 31:33 -- Weird behaviors from Claude 4 The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity. 

Claude 4’s system prompt leaked? In episode 57 of Mixture of Experts, host Tim Hwang is joined by Chris Hay, Kate Soule and Aaron Baughman to debrief a hectic week in AI.  First, Anthropic’s system prompt for Claude 4 was leaked; what stuck out to our experts? Then, Rick Rubin and Anthropic are vibe coding? We debrief “The Way of Code.” Next, OpenAI paid $6.5 billion for Jony Ive’s company, LoveFrom . Finally, Microsoft theorizes the development of “agent factories”. Is there a “winner takes all” in the AI agent's space?  Tune in to this week’s Mixture of Experts for more! 00:01 – Intro 00:51 -- Claude 4 system prompt 13:23 -- The Way of Code 23:03 -- Jony Ive and OpenAI 32:30 -- Microsoft’s “agent factory”  The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity. 

How long until Anthropic drops Claude 5.0? On today’s bonus episode of Mixture of Experts, guest host Bryan Casey is joined by Chris Hay, Marina Danilevsky and Shobhit Varshney to analyze the newly released Claude 4.0 family: Opus 4 and Sonnet 4. What do we know about the model architecture vs. What is speculation? In this special episode we talk Anthropic, OpenAI, Google and the rest of the competition in the AI race! Who will win? Tune-in to this bonus Mixture of Experts for more!  00:01 – Intro 00:32 – Claude 4.0  12:48 – OpenAI and Jony Ives 22:27 – Anthropic full-stack   The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity. 

Should you pay for Google’s AI Ultra subscription plan? In episode 56 of Mixture of Experts, host Tim Hwang is joined by Abraham Daniels, Gabe Goodhart and Marina Danilevsky to debrief the announcements from Google I/O 2025.  Next, RedHat dropped llm-d, a Kubernetes-native distributed inference serving stack; what is it and why does it matter? Then, we analyze Microsoft’s NLWeb: is everything becoming conversational? Finally, Stack Overflow has been on a decline. Is AI to blame? Find out more on this week’s Mixture of Experts! 00:01 – Intro 00:52 --Google I/O 2025 announcements 11:36 -- Stack Overflow 22:04 -- llm-d 30:08 -- NLWeb  The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity. 

Can Mistral make Europe a global AI contender? In episode 55 of Mixture of Experts, host Tim Hwang is joined by Chris Hay, Volkmar Uhlig and Kaoutar El Maghraoui to discuss the drop of Mistral Medium 3. Next, we analyze the AI chip sales both NVIDIA and AMD made to Saudi Arabia. Then, with IBM’s new ITBench and OpenAI’s HealthBench, we dive deeper into benchmarks for AI evaluation. Tune in to this week’s Mixture of Experts for more! 00:01 – Intro 00:47 -- Mistral Medium 3 12:26 -- AI chips to Saudi Arabia 21:21 -- AI evaluation benchmarks 31:47 -- Amazon's AI-generated pause ads  The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity. 

Has AI hallucination gotten out of control? In episode 54 of Mixture of Experts, host Tim Hwang is joined by Kate Soule, Skyler Speakman and Kaoutar El Maghraoui to analyze reasoning models and rising hallucinations. Next, as IBM Think 2025 wraps, the experts unpack the biggest highlights from IBM’s biggest show of the year: new AI agents, Ferraris and ... penguins? Then, OpenAI is making moves with its acquisition of Windsurf. What does this mean? Tune in to this week’s Mixture of Experts for more! 00:01 – Intro 01:12 – IBM Think 2025 09:27 – Reasoning models and hallucinations 19:23 – OpenAI Windsurf acquisition   The opinions expressed in this podcast are solely those of the participants and do not necessarily reflect the views of IBM or any other organization or entity.