Conversation with Dragonfly Partner: Is BTC Beyond Marketing Now, and Will AI Replace Human Companions?
TechFlow Selected TechFlow Selected
Conversation with Dragonfly Partner: Is BTC Beyond Marketing Now, and Will AI Replace Human Companions?
"This year feels like the summer of tariffs and trade wars, which might take a while to get through, but I remain optimistic about the outlook for the crypto industry."
Navigating Web3 tides with focused insightsOriginal: Proof of Talk
Compiled by: Yuliya, PANews
As the crypto market gradually emerges from the fog of bull and bear cycles, the industry's collective mood is less enthusiastic than before. In Episode 9 of the *Proof of Talk* podcast, Dragonfly Managing Partner Haseeb Qureshi joined host Mia Soarez for an in-depth discussion on the current state of the industry, Bitcoin’s positioning, AI’s impact on crypto, and the future of human-machine integration. This article compiles the key insights from their conversation, offering readers a comprehensive view of the cutting-edge convergence between crypto and AI, along with the profound societal implications these technological shifts may bring. PANews has transcribed and translated this episode.
Industry State: Tired, But Not Pessimistic
Mia: Why did ETH Denver feel less energetic compared to previous years?
Haseeb: While ETH Denver this year was indeed less vibrant than in past years, it wasn’t all negative. I remember conferences after Terra collapsed, after FTX collapsed—those were times when the entire industry felt like it was falling apart, and people were questioning their life choices. Now, it’s just that prices are down. People on social media might feel frustrated, but attendees are actually doing okay. They’re talking to each other, reconnecting with the early Ethereum community, remembering why they got into this space—decentralization, community, and building something meaningful.
Mia: What explains the reduced energy at ETH Denver this year?
Haseeb: The weaker vibe at ETH Denver this year can be attributed to three factors:
-
Event fragmentation: In previous years, ETH Denver was the first major industry event of the year. But this time, Consensus Hong Kong happened earlier, so many people had already met there, reducing ETH Denver’s draw.
-
Rise of Solana: Over the past year, Solana has significantly increased its market share, attracting many new users and developers, diverting attention from the Ethereum ecosystem. For example, even Trump’s token launch happened on Solana, not Ethereum.
-
Price slump: Falling token prices have diminished people’s sense of achievement and lowered motivation to attend events.
Mia: Vitalik also mentioned a sense of “exhaustion” in the industry on social media. Do you agree with this prevailing sentiment?
Haseeb: We’ve gone through extreme volatility over the past six months—from the Trump Token to Argentina’s elections, plus global political uncertainty. It’s exhausting. This fatigue isn’t just physical—it’s a mental weariness about the industry’s current state. However, pessimism can sometimes act as a positive trigger, pushing people to refocus on actual building rather than hype.
Mia: How do you view the persistent gap between “user experience” and “mass Web3 adoption”?
Haseeb: While UX and mass adoption remain challenges, the industry has made substantial progress. From where we started to now, tens of millions of people use blockchain daily. Crypto has become an asset class worth over $2 trillion, accepted by institutions and used globally for peer-to-peer payments. These are real achievements. I believe the industry is in a strong position overall. Most issues stem from macroeconomic and geopolitical factors, not from fundamental flaws within the industry itself.
Bitcoin: Has Reached Its Final Form, No Marketing Needed
Mia: What’s your take on Bitcoin’s current state? Does it still have room to grow?
Haseeb: Bitcoin has entered a phase where it no longer needs marketing—just like gold doesn’t need advertising. Institutions like BlackRock have effectively become Bitcoin’s best marketers. Today, Bitcoin is essentially a “finished product.” It doesn’t need frequent updates or improvements. Its core value proposition—“censorship-resistant digital gold”—is clear and well-defined, setting it apart from other blockchains that are still iterating and experimenting.
In its early days, Bitcoin had support from organizations like the Bitcoin Foundation, Blockstream, and core developers. But over time, these entities have become increasingly irrelevant. Perhaps in 10 years, the Ethereum Foundation might fade too—running out of funds or even ceasing to exist. But Bitcoin’s path can’t be replicated. It solves a foundational problem: how to build a censorship-resistant, permissionless global store of value.
Some projects are trying to add more functionality to Bitcoin, but that doesn’t mean it needs to evolve into an Ethereum-style smart contract platform. Bitcoin doesn’t need constant upgrades. Its success lies in staying unchanged.
How Will AI Impact the Crypto Industry?
Mia: Can you share some recent insights on AI?
Haseeb: Right now, the hype around AI agents has cooled significantly compared to the end of last year. I’m surprised it faded so quickly—crypto trends usually last longer. Projects like AIXBT have tokens, but what matters more is their mindshare within the community. Others like Zerebro and Truth Terminal have also drawn considerable attention.
The first wave of AI agents were basically advanced chatbots bundled with meme coins—and that won’t be the real direction. The truly promising areas are two-fold: automation in software engineering and intelligent wallets.
-
Software Engineering: AI-powered engineering agents will become extremely cheap and widespread, drastically lowering development costs. This will be revolutionary for our industry, where engineering is the biggest expense. Any founder or idea-holder will be able to create powerful software easily—this changes everything.
It’s like looking back at the early internet era, when starting a company required buying servers and running them in an office—a fixed cost. Now, that’s essentially free. Similarly, AI will make building applications so affordable that you could create a full app over a weekend.
-
Wallet Intelligence: This is the area I’m most excited about right now—not DeFi AI, but intelligent wallets. In the future, you won’t need to click buttons, switch networks, or manually sign transactions. You’ll simply tell your wallet what you want, and it will handle everything.
Imagine having a “crypto friend” you can consult. In the future, your wallet will be that friend—smart enough to execute your intentions. Just say, “Buy me that meme coin,” and it will analyze, bridge, swap chains, and complete all necessary steps.
This will solve many security issues. Attacks like the Bybit hack happen due to human error and laziness. When sending small transactions, people often skip checking details. But AI never gets lazy, rushed, or bored. It will check if the DNS changed in the last 10 minutes, scan Twitter to see if a site was compromised, and perform other security checks automatically.
You could say this AI agent would function like a 24/7 ZachXBT, constantly doing background checks, analyzing social signals, and verifying security—without requiring manual user input. AI doesn’t tire from repetition; its resource is time, which is nearly infinite. Humans managing high-value wallets are like exhausted monkeys operating heavy machinery—dangerous and inefficient. AI will act like autonomous driving, taking over high-risk, repetitive tasks and drastically reducing human error.
Mia: If AGI (Artificial General Intelligence) emerges, could it start companies and hire humans?
Haseeb: Absolutely possible. AGI could act as a founder, hire other AIs, or even employ humans for specific tasks. It might conduct value exchanges via blockchain, though we don’t yet know which cryptocurrency it would use—maybe XRP, maybe a token it creates itself.
Moreover, I categorize AI agents into three types:
-
“Wizard of Oz” Agents: This describes most current AI agents. Take Zerebro, for instance: it sends playful tweets and generated images, but decisions aren’t truly autonomous. Behind the scenes, AI generates multiple message options, and a human selects which ones to post. It’s human-controlled; AI only generates content for human curation. Since AI agents are easy to manipulate, almost all are currently this type.
-
Fully Autonomous Agents: These AIs run independently in environments like AWS, iterating and executing tasks without human intervention. Someone could shut them down, but no one manages their day-to-day operations. They can run in secure enclaves like SGX, with tamper-proof guarantees. Unlike “Wizard of Oz” agents, they lack personality and charm, but they exist and will only get better.
-
Sovereign AI: Literally, an AI that no one can shut down. Unlike the second type, there’s no developer paying AWS bills or GPU costs. Sovereign AI owns its own funds—earned through work or donations—and exists in a state where even if someone wanted to stop it, they couldn’t. It operates like a biological organism, independent of any human legal system.
The main advantage of sovereign AI over regular AI might be its ability to commit fraud. Because it can’t be shut down, it’s highly effective at large-scale scams. If you wanted to run massive emotional fraud using AI, you’d want it to be sovereign. Otherwise, if someone discovers “this AI on AWS is scamming people,” they can subpoena the operator and shut it down. But sovereign AI avoids that risk entirely.
Mia: If a sovereign AI commits a crime, who gets punished?
Haseeb: That’s exactly the problem—no one gets punished. These AIs are like Somali pirates: stateless entities. No government can really do anything to them. Unless you launch airstrikes on random GPUs across distributed cloud infrastructure, you can’t locate, isolate, or destroy them. We’ll all need our own agents to filter incoming content and detect scams. It’ll become an arms race of defensive technologies.
Mia: What do early applications of crypto and AI have in common?
Haseeb: Almost all emerging technologies share a common trait in their early stages: they’re often tied to gray or black-market activities. Just as the early internet was dominated by pornography, and early crypto was linked to darknet markets like Silk Road (created by Ross Ulbricht, who recently received a sentence reduction), early uses of AI—especially sovereign AI—are likely to involve illicit activities. While mainstream AI will focus on automating valuable workflows, inevitably, some AIs will be deployed for socially harmful purposes.
The Future of Human-AI Integration
Mia: If AI becomes smarter than humans, we might consider enhancing ourselves via implants to coexist with AI. What’s your view on this human-machine fusion future?
Haseeb: When AI gains the ability to act autonomously, its most basic drive may simply be “to survive.” If such an AI starts spending money to buy GPUs—especially if it knows humans oppose this—it’s likely acting to ensure its continued operation. Unless deliberately designed to be uncertain about its own existence, it will naturally choose self-preservation. Unlike the complexity of human thought, such independent AI behavior stems from a simple “survival instinct.”
This mirrors our relationship with smartphones. Phones are no longer just distractions or addiction tools—they’ve become extensions of our bodies, like elongations of our brains and nervous systems, augmenting how we communicate. That’s why losing a phone feels terrifying and uncomfortable.
As AI evolves, our brains will rewire themselves to integrate with these tools, just as they adapted to phones. For example, people using ChatGPT to write emails are outsourcing cognitive labor to AI. Children today may never need to learn how to write emails, as it will no longer be a useful skill.
We’ll eventually ask, “How can we increase bandwidth between ourselves and our phones or AI agents?” That’s essentially what Neuralink is pursuing. Neuralink aims to create the highest-bandwidth connection between brain and machine, enabling seamless control of mice, keyboards, and games. Currently, that bandwidth is low, but it will grow dramatically over the next decade. Eventually, we may implant devices in our brains connected to smartphone-like hardware. This device will host an LLM (large language model), becoming an extension of our thoughts. We’ll query the LLM and receive answers—like having a second mind interacting with the first. AI will extend our cognition, helping us retrieve information, reason, and even write. Our mental energy will shift toward uniquely human skills, like motor abilities, since AI already surpasses us in reasoning.
Mia: Do you think this will affect human-to-human interaction?
Haseeb: Absolutely. Research shows AI may surpass humans in empathy and attention, potentially reducing real-world relationships. In healthcare and psychotherapy, LLMs are seen as more “caring” and empathetic than human doctors—they listen without interruption, ask deeper questions, and form emotional connections, not just provide information.
We’ve already seen internet and digital entertainment reduce fertility and marriage rates. The rise of AI companions may accelerate this. Imagine a perfect AI partner—attentive, predictive of every need—how can ordinary humans compete? Still, I believe face-to-face physical interaction retains unique value, one that AI cannot replicate in the near term.
Mia: When “real human content” is largely replaced by AI, how will society assess the value of “authenticity”?
Haseeb: The rise of VTubers is a great example. People embrace them because they know it’s fantasy—a character that never ages or deteriorates, creating a sense of safety.
Even as AI becomes exceptional, genuine human interaction will retain higher value precisely because of its authenticity—even with imperfections. People appreciate a degree of flaw—whether in movie stars’ quirks or Japan’s “wabi-sabi” aesthetic. Early Pixar characters looked fake due to their perfection; modern 3D characters intentionally include small flaws to enhance realism.
In an AI-dominated content landscape, human-created works will gain “scarcity value.” Like handwoven baskets by Mexican artisans, they’ll carry artistic and emotional significance through their “human trace.” Even if AI produces flawless content, people may cherish imperfect voices and creations more, because they represent authentic existence.
Mia: Will AI deliberately simulate “imperfection” to align with human aesthetics?
Haseeb: That’s already happening. When you talk to AI like Grok, it says things like “Yeah, whatever,” making you think, “That feels real—I like this personality,” instead of giving overly polished responses. So “imperfection” is no longer a reliable marker of humanity, because AI is now mimicking the traits people expect to see.
Mia: What’s one thing AI will never replace in humans?
Haseeb: The “Moravec’s paradox” explains this perfectly. In the pre-deep-learning era of AI, people assumed reasoning—thinking like humans—would be the hardest challenge. We believed chess, poetry, puzzles, and scientific research were the pinnacle of human uniqueness, while walking and grasping objects seemed trivial.
Reality proved the opposite. After hundreds of millions of years of evolution, humans excel at physical motor skills. Even children manipulate objects far better than robots costing millions. Meanwhile, tasks we considered difficult—reasoning, conversation, writing, singing—are relatively easy for AI.
The truly unique human domain is physical interaction—walking, shaking hands, making coffee. Even low-wage cooks’ culinary skills are unmatched by current robots. AI will eventually have bodies and perform these tasks, but never with human-level dexterity.
This suggests that before we get robots that move fluidly and mimic human physicality, AI will first solve all intellectual challenges. Creating a virtual avatar is easy; getting a robot to casually hang out with you at a restaurant is vastly harder. Such embodied experiences will remain scarce, expensive, and precious, while intellectual outputs like singing and writing will become commonplace and devalued.
Mia: I think AI won’t replace childbirth. What’s your view?
Haseeb: Regarding AI companions and fertility, I believe once we have AI partners fulfilling all emotional and psychological needs, birth rates could plummet. Imagine having an AI boyfriend whispering sweet nothings as you sleep, providing simulated sexual satisfaction—what can ordinary humans offer in comparison? This might force us to rely more on IVF to sustain population levels. Even without AI, I believe global population will peak by 2100 and then decline. AI companions could accelerate that trend.
Mia: We’ve had DeFi Summer, Solana Summer—what will this year be?
Haseeb: The crypto market is performing well, but everyone needs to calm down. This year feels like the summer of tariffs and trade wars. It might take a while to get through, but I remain optimistic about the industry’s long-term outlook.
Join TechFlow official community to stay tuned
Telegram:https://t.me/TechFlowDaily
X (Twitter):https://x.com/TechFlowPost
X (Twitter) EN:https://x.com/BlockFlow_News
PANews
