DiscussionTechnical

How a Top AI Expert Actually Uses Claude

Marketing Against The Grain39m 14s

Ruben Hasid, a prominent Claude AI expert, discusses how AI should become invisible to users, the shift away from AI-generated writing toward authentic human thinking, and his practical Claude setup using memory and connectors instead of outdated folder systems. He argues that technology should evolve to eliminate user friction while emphasizing that human connection, social skills, and original thought remain irreplaceable in an AI-driven world.

Summary

In this episode of Marketing Against the Grain, host Kieran and guest Ruben Hasid explore the current state of AI adoption and practical Claude usage. Ruben begins by addressing a counterintuitive trend: despite his expertise in AI-assisted writing, he and other creators are now writing less with AI and more by hand, citing that writing is fundamentally an act of thinking. When AI generates content in a constrained style, it becomes predictable and boring, and users risk adopting AI's writing patterns rather than developing their own voice.

Ruben explains that early AI workflows relied on heavy context bloating through folders and file systems in tools like Obsidian, which was necessary when models had smaller context windows (3,000-4,000 tokens in 2022). He taught chain-of-thought prompting techniques and structured task breakdowns because models weren't capable of reasoning internally. However, modern models like Claude now have context windows of 1-10 million tokens and built-in reasoning capabilities, making these workarounds obsolete.

The core philosophy Ruben advocates is that "technology must become invisible." He criticizes the current practice of forcing users to choose between model tiers (Haiku, Sonnet, Opus, Claude 3.5) and effort levels, arguing this unnecessarily complicates the user experience. Future iterations of Claude should automatically route tasks to appropriate intelligence levels without user intervention. Similarly, the distinction between "coding" and "not coding" is artificial—Claude now performs code execution behind the scenes without users needing to understand this distinction.

Ruben walks through his actual Claude setup, highlighting key features: the unified chat and code interface, memory settings (which he sometimes disables for creative tasks to avoid constraining outputs), and connectors—particularly Gmail integration. He demonstrates how connecting Claude to his email and meeting notes allows him to draft emails that match his communication style without writing them manually, treating email composition like code: functional and replicable.

On the broader topic of AI agents, Ruben argues there's massive misconception around the term. He contends that users already interact with agents daily through Claude and ChatGPT; the technology is simply becoming more sophisticated and automated through routines and personal agent systems like Claude Duet. However, he expresses skepticism about lifestyle AI that automates shopping or booking flights, suggesting people actually enjoy these activities and don't seek further time-saving measures.

Regarding B2B transactions, Ruben predicts that while SMB and mid-market may see agent-to-agent transactions commoditized, high-value enterprise deals (six to seven figures) will always require human involvement and accountability. He contrasts this with the broader software industry, predicting that traditional SaaS companies will see their valuations compressed as AI-native alternatives emerge—companies that adapt will capture exponentially more value than legacy software.

The conversation shifts to the tension between AI empowerment and human isolation. Ruben shares a personal anecdote about using Claude to identify medication while sick, but then calling his mother anyway to maintain connection. He illustrates how technology-mediated solutions, while efficient, replace the human interaction and serendipitous conversations that actually matter. He presents cars and the internet as examples of technologies that are difficult to classify as purely good or bad—they enable tremendous progress but come with real downsides (deaths from car accidents, social fragmentation from digital communication).

On AI company accountability, Ruben expresses concern about regulatory capture as it happened with banks post-2008, where bailouts removed consequences for poor decision-making. He advocates for holding AI companies personally accountable through capitalism's natural incentive structure: if they build harmful products, the company suffers. He cites Anthropic's approach of stress-testing Claude before release to prevent misuse as an example of responsible development that shouldn't require heavy-handed regulation.

The hosts conclude by emphasizing that purely transactional, agent-to-agent interaction risks further isolating humans from meaningful connection. The warning is implicit: while AI dramatically amplifies individual capability, society must intentionally preserve human interaction, not allow technology to erode it through convenience.

About this episode

Framework + 6 Prompts to use Claude like an AI Expert: https://clickhubspot.com/bpsv Ep. 464 Is 95% of today’s writing influenced by AI? Kipp, Kieran, and guest Ruben Hassid (author of “How to AI” and founder of GPC) dive into how AI is transforming marketing, the future of software, and why human connection still matters more than ever. Learn more on how to master Claude with must-have settings, why writing is thinking (and why that matters in an AI world), and how the next era of enterprise marketing will blend invisible technology with personal connection for real business growth. Mentions Ruben Hassid https://www.linkedin.com/in/ruben-hassid/?isSelfProfile=false GPC https://www.gpc.consulting/ How To AI https://ruben.substack.com/ Loop: Outlearn. Outmarket. Outgrow https://www.hubspot.com/loop-marketing-book Claude https://claude.ai/ Grok https://grok.com/ Get our guide to build your own Custom GPT: https://clickhubspot.com/customgpt Resource [Free] Steal our favorite AI Prompts featured on the show! Grab them here: https://clickhubspot.com/aip We’re on Social Media! Follow us for everyday marketing wisdom straight to your feed YouTube: ​​https://www.youtube.com/@matgpod  Twitter: https://twitter.com/matgpod  TikTok: https://www.tiktok.com/@marketingatg Thank you for tuning into Marketing Against The Grain! Don’t forget to hit subscribe and follow us on Apple Podcasts (so you never miss an episode)! https://podcasts.apple.com/us/podcast/marketing-against-the-grain/id1616700934   We really appreciate your support. Host Links: Kipp Bodnar, https://twitter.com/kippbodnar   Kieran Flanagan, https://twitter.com/searchbrat  ‘Marketing Against The Grain’ is a HubSpot Original Podcast // Brought to you by Hubspot Media // Produced by Darren Clarke.

Key Insights

  • Ruben argues that using heavy context bloating with folders and files to constrain AI's writing style actually prevents the model from growing and evolving, making outputs predictable and boring over time.
  • He observes that modern AI models (like Claude with 1-10 million token context windows) have eliminated the technical constraints that previously necessitated complex workflow structures, yet users still cling to outdated practices.
  • Ruben claims that Claude now performs code execution and reasoning internally without users knowing or choosing it, meaning the distinction between 'coding' and 'not coding' is already functionally obsolete.
  • He predicts that AI companies shouldn't use model tiers (Haiku, Sonnet, Opus) in user-facing interfaces, arguing this adds cognitive friction that should be hidden by the system automatically routing tasks to appropriate intelligence levels.
  • Ruben contends that the term 'AI agents' is misleading because users already interact with agentic systems daily through Claude and ChatGPT without realizing it; the word creates false distinction rather than describing new capability.
  • He argues that traditional SaaS valuations will compress as AI-native alternatives emerge, not because they fail today but because markets price in the inevitability of disruption, similar to how legacy industries were revalued before they actually declined.
  • Ruben asserts that high-value B2B deals (six to seven figure contracts) will always require human involvement because trust, relationship, and accountability cannot be delegated to agent-to-agent transactions, even if lower-value transactions become automated.
  • He claims that corporate accountability through capitalism's natural incentive structure (companies suffer when products cause harm) is preferable to regulatory capture, using banking's post-2008 bailouts as an example of how regulation can remove consequences and enable poor behavior to continue.

Topics

Claude AI capabilities and best practicesThe evolution of AI model architecture and accessibilityTechnology becoming invisible versus visible complexityWriting as thinking and authenticity in content creationAI agents and misconceptions around AI automationMemory and connector features in Claude setupHuman connection versus AI-mediated efficiencyB2B transaction automation and legacy software disruptionCorporate accountability in AI developmentTechnology's societal trade-offs (benefits versus downsides)

Transcript

AI models have exploded. They can do more things than you could have ever dreamed about, but everyone is still struggling with their adoption. Like, how do I use this stuff? Well, we have the perfect guy to solve this problem. going to give you all of the secrets and you want to stay tuned because in this episode he opens up the screen and gives you his exact settings that you can just copy straight after this episode so you too can become an ai master all of that and more on this episode of marketing against the green all right we have the most knowledgeable expert on all things AI, and in particular, Claude. I follow all of…

Full transcript available for MurmurCast members

Sign Up to Access

More from Marketing Against The Grain

Get AI summaries like this delivered to your inbox daily

Get AI summaries delivered to your inbox

MurmurCast summarizes your YouTube channels, podcasts, and newsletters into one daily email digest.