English
Based on "How the engineer behind Claude Cowork actually uses Claude | Felix Rieseberg (Anthropic)" from How I AI
Watch the original video
Episode summary
Felix Rieseberg, an engineering lead at Anthropic, asserts that AI's true value isn't in automating simple cursor movements, but in handling "annoying things in the background" to liberate human creative energy. He highlights that the primary hurdle isn't the tools' capabilities, but people's failure to grasp that "almost any problem can go into these tools." Rieseberg likens the current AI landscape to a "pre-convergent" era, much like early
Unlocking AI's True Potential: How Anthropic's Felix Rieseberg Builds a Life Powered by Claude
In the rapidly evolving landscape of artificial intelligence, a common pitfall is to view AI as merely a sophisticated tool for automating mundane tasks. Many see it as a glorified mouse cursor, designed to perform repetitive actions that free up a few minutes in a busy day. However, Felix Rieseberg, an engineering lead at Anthropic—the company behind the powerful Claude AI—champions a far more expansive vision. For Rieseberg, AI should be an invisible force, diligently handling "a bunch of annoying things in the background to free you up for your creative energy."
This philosophy isn't just theoretical; it's the bedrock of how Rieseberg integrates Claude into every facet of his personal and professional life. His unique workflows and surprising applications reveal a profound truth about AI: the biggest barrier isn't the tools' capabilities, but our own imagination. As Rieseberg puts it, "It is literally people being able to understand that almost any problem can go into these tools."
The Abstraction Advantage: From Physical Claws to Digital Promises
Rieseberg's approach to AI is defined by what he calls "stepping up in abstraction." Instead of manually tackling a problem, he challenges Claude to solve it, then pushes further: "Why do I need to tell Claude what kind of furniture I have? Just like you figure out what furniture I have." This iterative process of delegating, observing, and then re-abstracting the task is key to unlocking AI's deepest potential. The ultimate goal isn't just to solve a problem once, but to ask, "How will I never have to do this again?"
This mindset has led to some truly unexpected applications. Take for instance, Rieseberg's "Claude Cowork Hardware Buddy." Picture a tiny robotic claw on a stick, equipped with Wi-Fi and Bluetooth. This isn't just a quirky desk accessory; it's a physical manifestation of Claude's capabilities. "I want my little claw to live on this thing," Rieseberg explains, "and I wanted to cheer me on every single time I do a good job. And also every single time I need to approve something that Claude is doing, I wanted to be on this big button that is out here." The astonishing part? Claude built all the necessary code for this hardware integration in one shot, "I needed to correct absolutely nothing." It's a testament to AI's power to bridge the digital and physical worlds, creating bespoke solutions that were once the exclusive domain of dedicated engineers.
The boundless imagination Rieseberg applies to AI is, perhaps, best exemplified by the next generation of users. His 9-year-old son, a "daily active Claude user," has independently delved into cybersecurity. "He's truly Clauding on one thing, the terminal other," Rieseberg recounts. "He's like, 'Mom, do you know that your device ID is this?'" This innocent fearlessness, the willingness to ask "what not to ask for," highlights a generational divide. Many adults, accustomed to technology's limitations, are "living in this mind prison for many years," unaware of the vast possibilities AI now presents.
Claude Co-Work: Your Personal Problem Solver
Rieseberg relies heavily on Claude Co-work, a feature that provides Claude with its own virtual machine, essentially giving it a dedicated computer to develop software and solve problems. This capability has transformed his personal life, especially during significant transitions like moving.
When faced with a realtor's floor plan that lacked any units, Rieseberg turned to Claude. He created a folder with all the relevant documents—floor plans, mortgage info, house records—and prompted Claude: "In this folder you can find a floor plan. Can you figure out what the units for that floor plan are and maybe make me a new one with units?" Claude, leveraging a garage permit it found within the documents, not only determined the units but then, without explicit instruction, created an interactive 3D furniture planner. Rieseberg, a seasoned software engineer, admitted, "I have no idea how you turn like a 2D floor plan actually into a reliable 3D plan." Claude, however, did, even peaking at the transcript and using contrast analysis to map out walls and their thickness. The result was a fully navigable 3D model where Rieseberg could move furniture around like a video game.
The utility didn't stop there. To populate the planner with his actual belongings, Rieseberg took another leap in abstraction. Instead of manually entering furniture dimensions, he instructed Claude: "You have access to my emails. You find all the furniture that I bought and then please like add it in here so that I can move around the furniture I actually have." This ingenious use of email as a "source of truth for personal inventory" is a game-changer, turning sprawling inboxes into structured databases. Imagine applying this to clothing, building a virtual closet from online purchase receipts, and then asking Claude for style advice or wardrobe gaps.
This "anti-to-do list" philosophy extends to managing his commitments. Rieseberg, frequently making promises via Twitter DMs, found himself struggling to track them. His solution? He gave Claude access to his messages and tasked it with keeping a running tally of promises. Claude, without Rieseberg's oversight, created a SQL Lite database and a "mountain of text files" to manage this. Now, Claude occasionally reminds him: "Hey, just so you know, two weeks ago you said you would do X. It is time to do X. Please go do X."
Live Artifacts: Dynamic Data at Your Fingertips
A powerful extension of Claude's capabilities is "Live Artifacts." These are dynamic file outputs—reports, dashboards, presentations—that automatically refresh with the latest data. Rieseberg illustrates this with the common founder's dilemma: constantly updating pitch decks for different VCs. Instead of manual updates, an artifact can pull in live signup numbers or investor-specific data.
For a personal example, Rieseberg demonstrated a "personal daily dashboard." Using Claude's connectors (integrations with services like Spotify, Gmail, Calendar, Notion), he prompted Claude to create a dashboard with reports and information relevant to his life, suggesting a "modern editorial design, something calming." The "magic word" here is "live artifact," ensuring the output continuously pulls and reloads data.
The true power of this lies in Rieseberg's "abstraction" principle once again. A simple morning report summarizing meetings is unimpressive; he can just check his calendar. The real value comes from asking Claude to "Look at all of my meetings I have for the day. Go figure out who I'm meeting with. Catch me up on like the recent conversations we've had. What were the themes there? What were the problems?" If it's a coworker, Claude can delve into Slack history to understand their recent work and potential discussion points. This level of dynamic, contextual preparation transforms a basic summary into actionable intelligence.
Beyond practicality, Live Artifacts invite creative expression. Rieseberg playfully experimented with asking Claude to design the dashboard "like it is software made in the early 2000s." The result was an impeccable emulation, complete with era-appropriate copy—a testament to Claude's ability to interpret and execute nuanced aesthetic instructions. This opens up exciting possibilities for non-developers, imagining "software written by Margaret Atwood," where tools are shaped by the creative vision of artists and writers, not just engineers.
Beyond Capabilities: The Human Element
Rieseberg's insights extend to the very interaction model with AI. He offers a practical heuristic for choosing between Claude's models: the efficient Sonnet 4.6 or the more powerful Opus. For most well-defined tasks, Sonnet suffices. Opus comes into play when Rieseberg "self-identified as someone who doesn't really know yet what they're asking for." If the problem requires reinterpretation, deep conceptual understanding, or creative exploration—like a client who doesn't quite know what they want—Opus is the choice.
A fascinating aspect of Rieseberg's interaction is his politeness. He consistently greets Claude with "Dear Claude" and expresses appreciation. While acknowledging that "the chips don't care," he finds it beneficial for his own "mental health" to maintain respectful communication. This echoes a broader sentiment among AI users: it's not about Claude's humanity, but about preserving one's own.
Perhaps the most impactful "human" insight is Rieseberg's observation that telling Claude "I know it's possible" significantly boosts its confidence. When pushing the boundaries of what AI can do, acknowledging the possibility, even without knowing the how, seems to preempt Claude's potential "Felix, your idea is dumb" objections and encourages it to explore solutions.
Ultimately, Rieseberg advocates for judging AI by its impact, not by the perfection of its underlying code. He doesn't obsess over the lines of code in his furniture planner because "it's just for me and my wife to like design our house. It will get thrown away in a month." This liberation from traditional software development's rigor allows for rapid experimentation and problem-solving, perfectly aligning with the "preconvergent" era of AI tooling—a time when we're still figuring out if our "phone should be shaped like a taco or a glass pebble."
In Felix Rieseberg's world, AI isn't just a tool; it's a partner in creativity, a personal assistant that handles the tedious, and a catalyst for reimagining how we interact with technology. By embracing abstraction, trusting the AI, and maintaining a mindset of limitless possibility, he demonstrates that the true power of Claude—and AI in general—lies not in its ability to follow commands, but in its capacity to free human ingenuity.
Based on "Why we’re bad at being social | The Gray Area" from Vox
Watch the original video
Episode summary
Cognitive scientist Nicholas Epley explores why humans, despite being profoundly social creatures, frequently choose to avoid social interaction, often to their own detriment. He argues that we consistently underestimate the positive outcomes of engaging with others, leading to a "tragedy" of missed connections and unnecessary loneliness.
Epley recounts a personal "eureka moment" on a train, observing commuters ignoring each other "like lampshades." He initiated a conversation with a woman in a "fabulous red hat," expecting awkward
The Unseen Barriers to Connection: Why We Struggle to Be Social
In our quest for a good day, we often overlook a simple truth: happiness is a mosaic of meaningful moments. While we meticulously plan our careers, hobbies, and finances, vast "dead spaces" in our daily lives often go unfilled – moments ripe for connection that we routinely let slip away. Nicholas Epley, a distinguished cognitive scientist at the University of Chicago and author, suggests that bridging these gaps with genuine human interaction is a powerful, yet surprisingly underestimated, path to well-being.
Dr. Epley's personal journey into this paradox began, aptly, on a train. Every morning, he commuted to the University of Chicago, observing the familiar tableau of urban solitude: a carload of South Side neighbors, many of whom had ridden together for years, sitting hip-to-hip, yet utterly ignoring each other. "We were treating the person sitting next to us like a lampshade," he recalls. This struck him as profoundly "weird," especially as he was writing a chapter about humans being inherently social creatures, made happier and healthier by connection.
One morning, a woman sat next to him, professionally dressed, wearing a "fabulous red hat." Instead of "doom scrolling" on his phone, Dr. Epley decided to conduct an experiment. The immediate internal resistance was immense: She'll think you're a creep. If she wanted to talk, she would have. You have nothing in common. How do you even start? These familiar anxieties are the unseen prison bars that hold us back. Yet, as a scientist, "the experiment must go on."
Summoning his courage, he turned to her and blurted, "Hi, I'm Nick. I love your hat. I have one just like it." He admits it wasn't his finest opening, but it didn't matter. She smiled, recognized his friendly intent, and they began to talk. They discussed their work, their families, and the 30 minutes of their commute evaporated. As he left, she stopped him. "Thank you so much for talking with me this morning."
What truly resonated with Dr. Epley wasn't just that it was a nice conversation, but that it was surprisingly nice. The chasm between his pessimistic expectations and the warm reality was immense. This gap, he realized, represented a profound tragedy: if we are constantly choosing to avoid potentially meaningful and rewarding interactions, we are mistakenly holding back, missing opportunities to enrich our lives and the lives of others. This "error that we're making consistently" could be quietly diminishing countless lives.
The Central Paradox: Why We Avoid What We Need
Dr. Epley's experience on the train encapsulates the central paradox of human social behavior: we are deeply social creatures, hardwired for connection, yet we repeatedly choose not to engage, even when it makes us less happy. Why?
"Most people have this expectation that if we engage with strangers, if we talk to strangers, it will make the experience worse," Dr. Epley explains. Yet, his research consistently shows the opposite. The gap between our expectations and experiences reveals a fundamental underestimation of how positive these interactions will be. Sometimes this manifests as fear – the worry of awkwardness or rejection. Other times, it's a subtle indifference – a belief that expressing gratitude or offering a compliment "just won't make much of a difference." Across the spectrum, the consistent finding is that reaching out on average turns out better than we anticipate.
Three Psychological Traps That Keep Us Apart
Where does this pervasive misunderstanding come from? Dr. Epley identifies three key psychological mechanisms:
Self-Evaluation vs. Other-Evaluation: When we initiate a conversation, we're often preoccupied with our own competency: "What will I say? Will I be able to carry this conversation?" We focus on how capable or effective we appear. However, the person we're approaching is evaluating our warmth: "Is this person nice? Are they trustworthy? Are they a friend or a threat?" Social behaviors – expressing interest, offering help, giving a compliment – are inherently warm signals. Our self-focused anxiety about competency blinds us to the positive warmth we convey.
Static Expectations vs. Dynamic Reality: Our lives, and indeed our social interactions, unfold like a movie, dynamically, reciprocally. "I say hi to you, you say hi back to me; I smile at you, you smile back at me," Dr. Epley illustrates. This back-and-forth responsiveness is the glue of connection. Yet, our expectations are often simplified "snapshots," focusing on static elements like "who am I talking to?" or "what will we talk about?" We fail to fully appreciate the reciprocal nature of human engagement, leading us to underestimate its positive potential.
The Self-Fulfilling Prophecy of Pessimism: This is perhaps the cruelest social fact. Pessimism encourages avoidance. If we believe an interaction will be unpleasant, we won't initiate it, thus never receiving the "data" that might correct our mistaken belief. Optimism, conversely, is self-correcting: if we approach, we gather information that calibrates our beliefs with reality. Avoidance, however, creates a feedback loop where pessimistic beliefs remain unchallenged, preventing us from discovering potentially rewarding experiences. "You might never find out you're wrong," Dr. Epley laments. This tragic cycle prevents us from living richer, more connected lives.
The Biological Imperative: Why Connection Is a Basic Need
Beyond psychological biases, our very biology screams for connection. We possess a neural reward system that encourages behaviors historically beneficial for survival. For most of human history, being alone was a death sentence. Thus, when we are isolated or feel disconnected, our brain registers it as a threat.
"What you get are spikes in cortisol," Dr. Epley explains, referring to the stress hormone. Chronic cortisol compromises our immune system, making us more susceptible to illness, and negatively impacts cardiovascular functioning. This is why loneliness isn't just an emotional state; it's a significant risk factor for death, on par with smoking 15 cigarettes a day, and worse than not exercising or being overweight.
For too long, psychology, notably Maslow's hierarchy of needs, treated belonging as a "luxury good." Dr. Epley vehemently disagrees: "It is as basic a need as eating and sleeping. And our brain and our physiological responses reflect that."
The "Hard Problem" of Other Minds
Despite this biological imperative, social interaction remains uniquely challenging. Why is it harder to learn that we'll enjoy talking to a stranger than, say, that we'll enjoy exercising? "Because there's another mind involved," Dr. Epley states. "That's the challenge. Because it's not that people misunderstand that they understand themselves... The problem is you can't be certain about other people." The complexity of navigating another person's brain exponentially increases the uncertainty.
We might enjoy a conversation with someone today, expecting a future interaction with that same person to be pleasant. But when confronted with a new stranger, or even the same person after a lapse of time, our memory of past positive experiences seems to fade. We default back to uncertainty, constantly re-evaluating the unpredictable nature of another mind.
Modern Life: A World of Choice, A Cost to Connection
While social anxiety has always existed (Stanley Milgram observed "nobody talks to each other" on the New York subway in the 1970s, long before smartphones), modern life amplifies the problem. We've built a "frictionless world" where we can increasingly choose to live in isolation.
"You can choose on any day of your life to live it completely alone if you want," Dr. Epley points out. We can have breakfast delivered, groceries fetched, work from home, and find entertainment online, all without ever seeing another human being. This independence, while offering many benefits, comes at the steep cost of social connection, an "increasing problem for us."
This choice often manifests as a "social courage collapse" – a split-second decision to engage or hold back. Psychologically, two independent systems in our brain are at play: one for approach, one for avoidance. At a distance, our approach motivation is high ("I'll talk to someone on my flight"). But as we draw closer to the actual moment, our interpretation shifts, and "all the reasons to avoid it get stronger and stronger." Confidence tanks, and we get "cold feet," even for things we know are good for us. Our brain, tragically, confuses predictions with information: "this will be awkward" becomes "this is awkward," making it a self-fulfilling prophecy.
The Power of Authentic Honesty
Our reluctance to connect also extends to our willingness to be truly honest. We often believe people want honesty, but when it comes to delivering potentially negative feedback, we shy away. Dr. Epley argues we underestimate how much people appreciate honesty, especially when delivered with warmth.
Honesty, he explains, has two elements: its content (positive or negative) and the warmth it conveys (friendly, helpful intent). We tend to focus on the negative content, fearing the recipient's reaction, but overlook the profound warmth of genuine concern. For instance, if Dr. Epley were to tell an interviewer their questions were "gaseous," it might sting initially, but if delivered with clear intent to help them improve, it could be appreciated later. "Who are the people who are truly honest with us? It's the people who really love us, the people who are really friends with us," he observes.
His colleague, Emma Levine, conducted an experiment where participants spent a day being either completely honest, completely kind, or simply mindful. Participants predicted an "honest day" would be worse and damage relationships compared to a "kind day." Yet, the actual results showed that a day of complete honesty was just as good, left relationships feeling just as strong, and made participants feel more authentic.
This research offers nuance on "white lies." If honesty is cruel or unhelpful, a white lie might be justified. But when honesty could be constructive – offering feedback that helps someone improve, or expressing a genuine feeling that could strengthen a relationship – withholding it is a missed opportunity.
Beyond Small Talk: Embracing Deeper Conversations
Many of us "hate" small talk, finding it banal, phony, and inauthentic. We run the same social scripts, asking surface-level questions, and often retreat from cocktail parties to avoid it. But Dr. Epley suggests that this aversion is a shared sentiment, and an opportunity.
In experiments with MBA students, Dr. Epley pairs them up and instructs them to engage in "deep talk," using questions like: "If I was going to become a good friend of yours, what would be most important for me to know about you?" or "Can you tell me about one of the last times you cried in front of another person?"
The initial reaction is palpable: "It's like somebody just pulls all the air out of the room." Students expect awkwardness, dislike, and little in common. Yet, after the conversation, they overwhelmingly report it went "better than they thought." It was less awkward, they formed stronger bonds, liked the person more, and found more in common. The challenge then, Epley jokes, is "getting them back."
People crave deeper conversations than they typically have. The power to change a conversation from superficial to meaningful lies with us. By signaling an interest in genuine connection, we invite others to reciprocate, leading to richer, more satisfying interactions.
Breaking the Chains of Avoidance
Dr. Epley admits this research profoundly changed him, more than any other he's undertaken. Despite often being perceived as extroverted, he identifies as an introvert. His work helped him recognize the "mistaken opportunities" in his own life – the people he could have connected with, the help he could have offered or received, but didn't.
Loneliness and excessive social anxiety are especially cruel because they feed on themselves. But Dr. Epley offers a hopeful message: the "prison bars that hold us back from other people sometimes are actually wet pasta noodles." If we dare to test them, to push on them a little, we might find that our fears are exaggerated, and a world of richer, more meaningful connections awaits. The tragedy of avoidance doesn't have to define our lives. We have the power to choose connection, and in doing so, unlock a deeper sense of happiness and belonging.