No Priors Ep. 91 | With Cohere Co-Founder and CEO Aidan Gomez

In this episode of No Priors, Sarah is joined by Aidan Gomez, cofounder and CEO of Cohere. Aidan reflects on his journey to co-authoring the groundbreaking 2017 paper, “Attention is All You Need,” during his internship, and shares his motivations for building Cohere, which delivers AI-powered language models and solutions for businesses. The discussion explores the current state of enterprise AI adoption and Aidan’s advice for companies navigating the build vs. buy decision for AI tools. They also examine the drivers behind the flattening of model improvements and discuss where large language models (LLMs) fall short for predictive tasks. The conversation explores what the market has yet to account for in the rapidly evolving AI ecosystem, as well as Aidan’s personal perspectives on AGI—what it might look like and when it could arrive. Sign up for new podcasts every week. Email feedback to show@no-priors.com Follow us on Twitter: @NoPriorsPod | @Saranormous | @EladGil | @AidanGomez Show Notes: 0:00 Introduction 0:36 Co-authoring “Attention is all you need” 2:27 Leaving Google and founding Cohere 4:04 Cohere’s mission and models 6:15 Pitfalls of current AI 8:14 How enterprises are deploying AI today 10:58 Build vs. buy strategy for AI tools 14:37 Barriers to enterprise adoption 20:04 Which types of companies should pretrain models? 24:25 Addressing flaws in open-source models 25:12 Current and expected progress in scaling laws 29:54 Advances in multi-step problem solving and reasoning 32:29 Key drivers behind the flattening curve of model improvements 36:25 Exploring AGI 39:59 Limitations of LLMs 42:10 What the market has mispriced

Sarah GuohostAidan GomezguestElad Gilhost

Nov 20, 202444mWatch on YouTube ↗

WHAT IT’S REALLY ABOUT

Cohere CEO Aidan Gomez on Enterprise AI, Reasoning, and Non-AGI Futures

Aidan Gomez, co-founder and CEO of Cohere and co-author of the Transformer paper, discusses how Cohere focuses on serving enterprises rather than competing for consumer chatbots. He explains the importance of robust foundation models, but emphasizes that enterprise success also depends on security, deployment flexibility, product structure, and helping customers avoid common implementation mistakes. Gomez outlines key use cases like RAG-based Q&A, summarization, and domain-specific assistants, and argues that reasoning-focused models and inference-time scaling will structurally change how AI capability is delivered and priced. He is skeptical of imminent AGI takeoff narratives, instead seeing a long, practical refactor of the economy using already-powerful but imperfect models, with model commoditization overstated and specialized model builders retaining leverage.

IDEAS WORTH REMEMBERING

5 ideas

Enterprises should start with simple customization before touching pre-training.

Gomez recommends a gradient of specialization: begin with fine-tuning and prompting changes, then move to post-training (SFT/RLHF), and only consider continuation pre-training for very large organizations with massive proprietary datasets and stringent performance needs.

Most failed enterprise AI POCs stem from RAG and prompting details, not model limits.

Cohere repeatedly sees failures because teams mis-format retrieved context, store data poorly, or assume models are human-like; structured APIs and more robust models can greatly reduce these failures.

Focus in-house efforts on AI systems that deliver unique competitive advantage.

Gomez advises enterprises to buy generic tools (e.g., broad copilots) and only build internally where the use case is specific to their business, like an insurance research assistant that shortens RFP cycles and directly drives revenue.

Security, privacy, and deployment flexibility are decisive for regulated industries.

In healthcare and finance, data often cannot leave a specific VPC or on-prem environment; Cohere’s ability to deploy in multiple environments is framed as a key differentiator and a prerequisite for accessing the most sensitive, valuable data.

Reasoning models shift improvement from pure training capex to inference-time spend.

Instead of waiting months for a new larger model, customers can pay for more inference-time compute to get smarter behavior on demand, changing both pricing models and infrastructure design priorities across the stack.

WORDS WORTH SAVING

5 quotes

We’re not going to build a ChatGPT competitor. What we want to build is a platform and a series of products to enable enterprises to adopt this technology and make it valuable.

— Aidan Gomez

People overestimate the models. They think they’re like humans, and that has led to a lot of repeat failures.

— Aidan Gomez

Even if we didn’t train a single new language model, there’s a half decade of work to go integrate this into the economy.

— Aidan Gomez

We’re pretty far along. We’re certainly past the point where if you just interact with a model, you can know how smart it is.

— Aidan Gomez

There’s a total technological refactor that’s going on right now and will last the next 10 to 15 years, and it’s kind of like we have to repave every road on the planet, and there’s four or five companies that know how to make concrete.

— Aidan Gomez

Aidan Gomez’s background and path from Google Brain to founding CohereCohere’s enterprise-focused mission, business model, and deployment strategyCommon enterprise pitfalls with LLMs, especially in RAG and promptingHigh-impact enterprise use cases: Q&A, summarization, research assistants, healthcare and insurance examplesTechnical strategy: fine-tuning vs post-training vs continuation pre-training for large enterprisesScaling laws, slowdown at the frontier, and the rise of reasoning and inference-time computeViews on AGI, superintelligence narratives, and whether models are becoming commoditized

High quality AI-generated summary created from speaker-labeled transcript.

Get more out of YouTube videos.

High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.