Podcast · Tech & Cybersécurité
Genie AI combines applied machine learning research with practical AI product development, positioning its leadership at the intersection of research and commercial implementation.
Using AI unpacks how machine learning researchers, founders, and growth professionals apply cutting-edge AI tools daily, moving beyond theory into implementation. The podcast examines competitive dynamics between OpenAI, Anthropic, Meta, and other AI companies, evaluating their technical approaches and business strategies. Episodes reveal which tools—from ChatGPT and Perplexity to Llama 2 and Mistral—solve real professional problems versus marketing hype.
Discover the latest episodes of Using AI to understand how artificial intelligence reshapes professional workflows.
Using AI moves beyond surface-level AI trends to examine how specific tools address real business problems. The podcast deconstructs the technical and strategic differences between competing platforms—why Anthropic emphasizes Constitutional AI while OpenAI prioritizes user experience, how Meta's Llama 2 democratizes LLM access, and what Perplexity and Nemo offer that OpenAI doesn't.
Episodes reveal that AI adoption isn't monolithic. A founder evaluates models differently than a research scientist. A growth marketer's needs differ from a product manager's. Using AI captures these distinctions, showing when GPT-4 becomes necessary versus when open-source alternatives like Mistral provide better ROI.
The podcast also covers emerging use cases: voice AI, multilingual translation systems, code generation for non-developers, and product management workflows enhanced by language models. This specificity means listeners gain actionable intelligence rather than generic "AI is transforming everything" messaging.
ML Research Scientists and Data Engineers benefit from technical deep dives into model architectures, transformer mechanics, and competitive positioning between research labs. Using AI addresses how Anthropic's approach to interpretability differs from OpenAI's scaling philosophy and what that means for research direction.
Founders and Growth Marketers gain clarity on which AI tools solve their specific problems—whether that's customer support automation, content generation, or code acceleration. The podcast evaluates pricing, integration friction, and time-to-value across competing platforms rather than assuming one tool fits all use cases.
Product Managers and Non-Technical Leaders learn how to evaluate AI without deep technical knowledge, understanding the business implications of choosing between OpenAI's proprietary models, Anthropic's safety focus, or Meta's open-source approach. Recent episodes on product management plus AI show how to integrate language models into product roadmaps effectively.
The podcast's episode titles expose a pattern: systematic competitive analysis. Rather than celebrating "AI innovation," Using AI repeatedly compares players—OpenAI versus Anthropic, Google versus OpenAI, Cohere versus OpenAI, Hugging Face versus OpenAI, Llama 2 versus OpenAI, Mistral versus OpenAI, Perplexity versus OpenAI.
This structure reveals that OpenAI functions as the industry reference point. Competitors position themselves by differentiation—Anthropic through safety guarantees, Meta through open-source accessibility, Cohere and Mistral through specialized capabilities, Perplexity through search-integrated answers.
Episodes also show that AI tooling has become granular. Rather than monolithic platforms, professionals now layer specialized tools: ChatGPT for general tasks, Code-Interpreter for development, voice systems for communication, translation models for multilingual work, and task-specific models like those from Hugging Face for edge cases. Using AI maps these combinations rather than assuming consolidation.
Understanding AI competitive dynamics changes how professionals evaluate tools. Rather than adopting the "market leader" by default, listeners can assess OpenAI, Anthropic, Meta, and others against specific criteria: safety guarantees matter if compliance is critical, open-source options reduce vendor lock-in, specialized models outperform general-purpose tools for narrow domains.
The podcast's technical discussions—transformer models, NLG approaches, AGI timelines—shift decision-making from emotion to architecture. Knowing that Anthropic's Constitutional AI prioritizes interpretability while OpenAI emphasizes scale changes how you evaluate cost-benefit tradeoffs.
For non-technical professionals, Using AI removes the barrier to AI adoption. Learning that Code-Interpreter enables product managers to prototype without engineers, or that voice systems eliminate translation friction, empowers immediate action. The podcast legitimizes using AI without mastering its internals.
Explore all Using AI episodes to deepen your AI strategy.
Listen to Using AI and stay ahead of enterprise AI shifts.
Transformer models use self-attention mechanisms to process and understand sequences of text by weighing the importance of different words in relation to each other. This architecture allows models like GPT-4 and Claude to generate coherent, contextually relevant responses. Transformers have become the foundation for nearly all modern large language models because they scale efficiently and capture long-range dependencies in text.
OpenAI leads with ChatGPT and GPT-4, prioritizing accessible consumer applications. Anthropic focuses on safety and interpretability through Constitutional AI. Meta emphasizes open-source development with Llama 2, enabling broader developer adoption. Each company pursues different strategic positioning: OpenAI dominates enterprise adoption, Anthropic attracts safety-conscious users, and Meta builds ecosystem loyalty through open tools.
Modern LLMs like GPT-4 and Mistral now handle multilingual content with near-human accuracy, reducing traditional machine translation friction. Specialized models from Cohere and emerging providers address specific language pairs and technical domains. Real-time voice translation through tools like Nemo represents a shift from text-only processing, enabling direct cross-language communication without intermediate steps.
Yes. AI coding tools like Code-Interpreter and similar systems abstract away syntax complexity, allowing product managers and growth marketers to prototype and deploy solutions without deep programming knowledge. These tools democratize development by translating intent into working code, though understanding fundamental logic principles remains valuable for quality control and iteration.
3 questions answered · 1 episode indexed
Rafi Farooq uses AI to draft user stories, behavior scenarios, and analytics event tracking by starting with a base prompt containing detailed company and product descriptions.
Galileo AI is a tool that generates high-fidelity wireframes through a chat interface where you input requirements and answer follow-up questions. The tool produces multiple wireframe options comparable to designs found on Dribbble, and crucially, allows you to import generated wireframes directly into Figma…
Dovetail is working on an AI feature specifically designed to tag user interviews and draw insights automatically. This is valuable because AI tagging reduces human bias in the tagging process compared to manual PM tagging.