One of Silicon Valley's most influential VC firms just cut most of its recruiting team. Andreessen Horowitz (@a16z) reportedly laid off the majority of its 34-person talent team, which supported hiring across its portfolio companies. At a time when exceptional talent has never been more critical to startup success. But perhaps the real story isn't about layoffs. It's about the economics of recruitment. For decades, recruiting has relied on centralized teams, proprietary databases and expensive intermediaries. A model built around maintaining permanent capacity, even when hiring demand fluctuates. What if the competitive advantage is shifting from owning recruiting resources to orchestrating access to the right talent networks? Imagine a distributed network of trusted professionals who identify, recommend and assess candidates based on their actual expertise, rather than relying exclusively on traditional recruiters. A network where contributors are rewarded for the value they create, not simply for maintaining an organizational structure. This is one of the hypotheses we're testing at @GoodHiveLabs, starting with Web3 talent. But decentralization alone solves nothing. Trust must be earned. Recommendations must be credible. Quality must be verified. And incentives must align the interests of candidates, companies and contributors. A distributed recruiting network without these mechanisms is just another database with a community label. The future may well be hybrid: smaller centralized teams coordinating broader networks of specialized contributors. Perhaps the next competitive advantage in recruitment won't be the size of your recruiting team, but the quality of the network you can activate. Can community-driven recruiting outperform traditional models? Source: Axios, October 14, 2026.

AI makes me feel superhuman. And that's exactly what scares me. Over a year ago, I used AI to draft complex contracts for @GoodHiveLabs, my collaborative Web3 recruitment platform. Agreements so specific that conventional templates were useless. Then came prospecting. Researching a company, understanding its business and crafting genuinely personalized outreach used to take hours per prospect. Scaling that was almost impossible. AI changed everything. A few months ago, I was spending three hours a day supervising AI workflows, checking results, testing and correcting errors. Today? Roughly ten minutes. I'm automating the supervision itself. And recently, another revelation: French accounting, taxes and administrative compliance. A bureaucratic nightmare I once thought required an ENA degree just to understand. Now AI helps me navigate it, ask better questions and understand my obligations. I've never felt so capable. So autonomous. So augmented. But here's the paradox. The more powerful AI makes me, the less human work my business seems to require. And I'm just one entrepreneur. Multiply that by millions of companies. I believe we're approaching a JOB APOCALYPSE. Call centers, administrative work, accounting and entire categories of traditional jobs could shrink dramatically over the next five years. Not everyone can reinvent themselves overnight. What happens to someone who has spent 30 years mastering a profession that AI can suddenly perform in seconds? We may need to rethink the very meaning of work, how productivity gains are distributed and whether universal basic income becomes a necessity. Are our political leaders remotely prepared? The AI revolution won't wait for society to catch up. #AI #FutureOfWork #JobApocalypse

Web3 doesn't just need more Solidity developers. It needs great engineers. That's something I've noticed in conversations with founders and CTOs hiring for Web3. They're not necessarily looking for someone who's spent five years in crypto. They're looking for strong engineering fundamentals, problem-solving skills, code quality and the ability to learn. And that changes how we should think about the Web3 talent shortage. Backend engineers, cybersecurity specialists, DevOps experts and distributed systems developers already possess many of the skills this industry needs. Of course, blockchain introduces specific challenges. Security, decentralization and irreversible transactions require specialized knowledge. For some roles, deep Web3 experience is essential. But an excellent backend engineer doesn't suddenly become a junior engineer because they're new to blockchain. At GoodHive, we believe the opportunity goes beyond competing for the same limited pool of experienced Web3 developers. It's also about recognizing transferable expertise and connecting great engineers with companies building decentralized infrastructure. For founders, that means looking beyond traditional Web3 profiles. For engineers, it means realizing that the skills they've spent years developing may already be highly valuable in this industry. The next generation of Web3 engineers may already be out there. They just don't call themselves Web3 developers yet. #Web3 #Blockchain #Engineering #TechHiring

€3 million sent to Mexico. Six months for the banks to figure out where the money had gone. Not a crypto hack. A real cross-border payment involving a CAC 40 company and major international banks. That story, shared today at @essec, probably says more about the future of blockchain than a thousand discussions about crypto prices. We gathered Julien Achard (@openbanq), @alixbouxaguet (@krakenfx_FR ), @pablo_veyrat (@merkl_xyz ) and Alexandre Mangin (@Team1FR ), moderated by Hayat Outahar (@GoodHiveLabs), around one question: What will the financial rails of tomorrow look like? Three takeaways: 1. Blockchain may not change what finance does. It changes how finance does it. Atomic settlement, 24/7 markets and smart contracts can remove layers of friction and delay. 2. This increasingly looks like convergence, not a war between TradFi and DeFi. Pablo made a great point: “DeFi” may actually be a bad name. On-chain finance or open finance better describes where we're heading. Julien made another provocative prediction: “By 2035, the RIB could be gone.” 3. The real battle may be about standards, currencies and network effects. Dollar stablecoins already dominate liquidity. Could decentralized financial rails actually reinforce the dollar’s global dominance rather than weaken it? The question is no longer simply whether finance will move on-chain. It’s which rails, standards, currencies and actors will shape it once it does. Great discussion with @KRYPTOSPHERE_ ESSEC, and proud to have contributed through GoodHive and the ESSEC Alumni Digital & Technology Club. Next stop: stablecoins and the future of money on December 3 - stay tuned.

What if blockchain doesn’t just improve finance, but gradually decentralizes it? Financial infrastructure is already digital. Yet it remains fragmented, dependent on multiple intermediaries, legacy systems and market structures designed decades ago. Blockchain introduces a different set of rails: programmable assets, 24/7 markets, on-chain settlement, DeFi and shared financial infrastructure. The first transformation is about efficiency. But the more interesting question comes next: What happens when traditional financial institutions start adopting decentralized rails? Does blockchain simply make the existing financial system more efficient? Or does it progressively change the role of banks, exchanges, protocols and market infrastructures themselves? That’s what we’ll discuss at ESSEC on October 7 with Julien Achard (@openbanq ), @alixbouxaguet (@krakenfx ), @pablo_veyrat (@merkl_xyz) and Alexandre Mangin (@AvaxTeam1), moderated by Hayat Outahar (GoodHive). 📍 ESSEC Cergy 📅 October 7 — 7 PM 🎟️ Career Days registration luma.com/l6bc839e Proud, through @GoodHiveLabs and the ESSEC Alumni Digital & Technology Club, to support @KRYPTOSPHERE_ ESSEC in bringing together the people building the next generation of financial infrastructure. Better financial rails, or a more decentralized financial system? Probably both. #Blockchain #DeFi #Tokenization #OnChain #GoodHive

Which skills should you develop to position yourself in Web3? Introducing @GoodHiveLabs first Web3 Talent Barometer. Each month, we’ll examine technical hiring requirements to help talent make informed learning and career decisions, and give employers a clearer view of the expertise teams need. Our first expanded research watchlist contains 121 technical roles across 58 companies. We classified 49 as backend or infrastructure, compared with 10 focused on smart contracts and protocol applications. The interesting part is how skills connect to the products being built. Among Web3-native teams, Aave’s vaults backend role combines indexing, orchestration, PostgreSQL and GraphQL, with Rust preferred. Stellar pairs C++ and Rust with consensus, peer-to-peer systems and performance. Coinflow connects application-stack knowledge with security engineering in payments. Institutional teams bring other combinations. Goldman Sachs connects Canton/Daml with EVM interoperability and tokenization. Wells Fargo names Go, consensus engineering, MPC and confidential computing. And the boundary between these groups is not clean. Crypto-native providers such as Anchorage Digital, Digital Asset and BitGo also build infrastructure serving institutions. For talent, our first takeaway is to connect your learning to the systems you want to build. DeFi applications, blockchain protocols, payments and custody require different combinations of expertise. For employers, a useful hiring brief should describe what an engineer must build, secure and operate, alongside the technologies involved. This first corpus is partial, with mixed signal dates. It is not a representative market sample or evidence of growth. We’ll progressively strengthen the methodology and coverage. Monthly observations should help us distinguish recurring requirements from genuinely emerging needs. Follow the series on the GoodHive blog: goodhive.io/blog First observations today. A clearer picture, edition by edition.

Stablecoins could weaken monetary policy. But follow the mechanism and you get to a much bigger question: how fast can we rebuild our monetary system without destabilizing the one we still depend on? Petra Tschudin from the Swiss National Bank recently discussed how new forms of money could affect central banks and monetary policy. It brought me back to a question that has bothered me for some time: Why aren't stablecoins allowed to pay interest? Under MiCA, stablecoin issuers can't simply pay holders interest. One reason is to prevent them from becoming too attractive as a store of value and competing directly with bank deposits. That's where it gets interesting. Banks can hold reserves at central banks and earn interest. Individuals can't. A liquid, programmable stablecoin passing the yield on its reserves back to holders could become a powerful competitor to bank deposits. So part of what regulation protects is an existing banking architecture our economies still depend on. Move money out of deposits too quickly and you risk disrupting bank funding, credit creation and monetary policy transmission. Yet regulation is also shaping the new infrastructure. The GENIUS Act and MiCA constrain how stablecoin reserves are held and managed. Monetary policy may not disappear in a tokenized economy. Its transmission mechanisms may evolve. And that's the real challenge: Move too fast, and you risk destabilizing the financial system. Move too slowly, and you protect yesterday's infrastructure while others build tomorrow's. The speed of this transition may become a competitive advantage for entire economies. We'll be discussing exactly these questions on December 3 in Paris at our ESSEC Alumni conference: Who will control the money of tomorrow?

I had a strange moment on LinkedIn this week: I couldn't tell whether I was talking to someone or to their AI. And it made me realize that the problem isn't AI helping us write. It's what happens when we stop reading what is being written in our own name. I use ChatGPT every day. Including to help me answer comments. But I read the comment. I think about it. Sometimes I disagree with the answer AI proposes. Sometimes we spend 10 minutes discussing it before I know what I actually want to say. The final words still have to reflect what I think. What's changing is that AI agents can now do the whole thing automatically. Follow prospects. Read their posts. Comment. Reply. Build familiarity. “Warm up” the relationship. Potentially without the human reading any of it. So imagine: Your AI agent comments on my post. I ask my AI to help me answer. Your agent reads the answer and replies again. We can now build a professional “relationship” while one of us, or perhaps both of us, isn't really participating in the conversation. And then this morning, almost perfectly timed, I saw LinkedIn's latest announcement about fighting automated engagement and “AI slop.” I don't think AI-generated writing is the problem. AI helping humans communicate is one thing. AI communicating in our name without us paying attention is another. Because automated engagement works by imitating a very human signal: I read what you wrote. I thought about it. I chose to respond. If that becomes automated at scale, the signal loses its meaning. And there's an even deeper issue. If my agent comments on something I haven't read, expressing an opinion I haven't considered, is that still my voice? What happens when it says something I don't actually believe? The more we automate human attention, the more valuable real human attention may become. In the agentic internet, the question may no longer be whether there's a human behind an account. It may be whether that human actually knows what their AI is saying.

What if decentralization doesn’t require everyone to do the work? Vitalik Buterin just outlined a long-term vision for Ethereum as a “cryptographic world computer.” The interesting part isn’t energy. Proof of Stake already reduced Ethereum’s energy consumption by ~99.95%. The real issue is scaling. Today, independent nodes still re-execute transactions to verify Ethereum’s state. If the network handles 10x or 100x more computation, validators eventually need increasingly powerful machines. At some point, scaling starts working against decentralization. Recursive STARKs could change the equation: Today: Execute → Execute → Execute → Execute Future: Execute → Prove → Verify → Verify → Verify A specialized prover performs the heavy computation and generates a cryptographic proof. Validators don’t need to trust the prover. They verify the proof. Proof of Stake doesn’t disappear. Validators still stake ETH and participate in consensus. What changes is how execution can be verified. This could allow Ethereum to scale without turning every validator into a data center. But there’s a paradox. Generating proofs is computationally expensive. What happens if proving becomes concentrated among a handful of specialized operators? Ethereum could centralize expensive computation while making independent verification dramatically more accessible. Which raises a deeper question: What actually needs to be distributed for a system to be decentralized? Maybe decentralization doesn’t require everyone to do the work. Maybe it requires everyone to be able to verify it.

A nice Sunday surprise: The Age of Giants has just been published by @hackernoon . It started with a question I've been thinking about for a while: What happens when AI gives ordinary humans capabilities that would have looked almost godlike a generation ago? Not another piece about AI replacing jobs. This one is about augmentation, power, responsibility, and what we choose to become with it. Thanks to HackerNoon for publishing it. hackernoon.com/the-age-of-gi…

Ethereum spent years scaling by moving activity away from L1. Now it's scaling L1 again. Glamsterdam, Ethereum's next major upgrade, is entering its public testing phase. One of the most striking signals: the latest devnet pushed the gas limit from 60M to 200M. But this isn't just about making blocks bigger. Ethereum is redesigning how blocks are built and executed, notably through ePBS and Block-Level Access Lists, while repricing expensive state operations to make higher capacity sustainable. The interesting part is the change in philosophy. Ethereum's scaling story used to sound simple: L1 for security, L2s for scale. The future looks more ambitious: L2s for massive scale. And a much more powerful L1 underneath them. Ethereum isn't finished infrastructure. It's still being rebuilt while running.

The ECB wants tokenized settlement to eventually run 24/7. Crypto has been doing that for years. But maybe that's precisely the point. With Pontes, the ECB is bringing DLT-based markets into settlement in central bank money, with plans to progressively extend operating hours toward 24/7. Of course, this isn't decentralization. And central bank settlement and DeFi don't solve the same problems. But something deeper is happening. For years, decentralization was supposed to replace financial institutions. Instead, institutions are starting to reproduce some of the properties decentralized systems made possible: programmable assets, more direct settlement, interoperability and always-on markets. Maybe decentralization's biggest victory won't be replacing institutions. Maybe it will be forcing institutions to redesign themselves.

Want to know where technology might be going? Follow the hiring. Apple is looking for expertise in stablecoins, tokenized deposits and blockchain for Apple Pay. Google Cloud is looking for RWA tokenization, stablecoin rails, custody and institutional Web3 infrastructure. Two jobs don't make a trend. And neither company is new to crypto. But what's interesting is how specific the skills are becoming. Apple is looking at the payment layer. Google at the infrastructure layer. Stablecoins. Tokenized deposits. RWA. Custody. Technologies pioneered inside Web3 are gradually finding their way into mainstream financial infrastructure. Product announcements tell us what companies are ready to show. Hiring can sometimes give us an earlier glimpse of what they are preparing to build. Two jobs don't make a trend. But they may tell us where to keep looking.

Crypto isn't entering traditional finance anymore. Traditional finance is moving on-chain. The SEC just opened the door to tokenized US stocks. But this is much bigger than the SEC or Wall Street. Look at 2026: • JPMorgan brought its dollar deposit token to Base. • DTCC used tokenized securities in real production transactions with 30+ major institutions. • BlackRock brought institutional money-market fund shares on-chain. • India issued ₹10.25B in corporate bonds on DLT, with settlement using the RBI's wholesale CBDC. • Circle launched Arc with BlackRock, DTCC, ICE, Mastercard, Standard Chartered and Visa among its founding validators. Europe is moving too: Euronext, CACEIS, LSEG, Deutsche Börse. And while mainland China remains tightly controlled, Hong Kong is becoming a major laboratory for tokenized finance. Different markets. Different regulations. Same direction. Money. Funds. Bonds. Collateral. Settlement. Market infrastructure. The global financial stack is moving on-chain. And we're seeing the same transformation through hiring at GoodHive: our estimates suggest that around 80% of the Web3 hiring signals we currently track come from institutional players or companies building digital asset infrastructure for them. That matters. Because putting financial markets on-chain isn't primarily a token problem. It's an engineering problem. Institutions need people who understand both worlds: blockchain + market infrastructure, smart contracts + settlement, crypto-native architecture + institutional constraints. Those profiles are still rare. The transformation has started. Now institutions need the people who can build it.

Circle launched Arc yesterday. And I keep coming back to one question: Why do we need another L1? Ethereum is secure, permissionless, EVM-native, and still home to much of the innovation, developers and liquidity in DeFi. Arc is EVM-compatible, but optimized for institutional finance: USDC gas, sub-second finality, privacy features and a permissioned validator set including BlackRock, Visa, Mastercard, DTCC and Standard Chartered. I understand the specialization. What I find harder to understand is why this needs to be a new L1 rather than a specialized L2 anchored to Ethereum. Because that choice raises a deeper question. One of blockchain's most powerful ideas was removing the need for financial participants to trust each other. Everyone plays by the same protocol rules. Arc partly puts trusted institutions back into the infrastructure itself. There may be a good reason for that. Financial systems sometimes need human discretion. Circuit breakers, emergency liquidity and coordinated intervention can stop a problem from becoming a systemic crisis. But that same discretion creates governance and conflict-of-interest risks. Financial history, from LIBOR to FX manipulation cases, gives us reasons to take those risks seriously. So how do we preserve enough human discretion to protect financial stability without recreating the concentration of power blockchain gave us an opportunity to rethink? As a strong believer in Web3, I welcome Circle's initiative. Bringing more of the global financial system on-chain is an important step forward. But I hope we don't just build a faster and more efficient financial system. I hope we build a more open and equitable one too. The technology matters. Who gets to control it matters even more.

The CLARITY Act failed yesterday. But some of its ideas shouldn't fail with it. Beyond the politics of the vote, what interests me is what CLARITY was trying to solve. One idea in particular: progressive decentralization. Most Web3 projects don't start decentralized. They start with founders, investors and a small team controlling most decisions. But if they succeed, ownership, governance and control can progressively move toward a broader network. CLARITY tried to recognize this evolution through the concept of a “mature blockchain system.” And that raises a fundamental question: How do you regulate an organization whose nature can fundamentally change over time? This matters for both sides of the ecosystem. Large financial institutions moving assets on-chain need regulatory certainty, standards and compliance. Startups need clarity too. But they also need proportionality, room to experiment and a regulatory path that can evolve with them. Europe already has MiCA. The US is experimenting with different approaches. Neither has all the answers. We should be learning from both. Financial infrastructure is changing. The organizations building it are changing too. Regulation needs to learn how to evolve with them.

Tokenization is only the beginning. Putting assets on-chain is relatively easy. Rebuilding the financial infrastructure around them is much harder. @KaikoData is an interesting example of where this is going. From its roots in crypto market data, it has expanded into pricing, benchmarks, indices and on-chain infrastructure, including through acquisitions such as Cometh. Now S&P Global is leading an investment extending its Series B to $110M, alongside players including @Nasdaq , BNP Paribas and RBC. Because tokenized markets still need pricing, liquidity, risk, smart contracts, execution, compliance and reliable data. Blockchain doesn't eliminate financial market infrastructure. It forces us to rebuild it. And increasingly, that infrastructure is being built now.

Europe wrote rules for crypto companies. America is starting to write rules for companies that may eventually stop being companies. That’s what I find most interesting about the CLARITY Act. A crypto project can start with 5 founders controlling the code, capital and governance, then progressively evolve into an open, community-governed, decentralized network. Regulation should be able to evolve with it. The US has already experimented with pragmatic structures like Wyoming’s DAO LLC. CLARITY could take another step by recognizing “mature blockchain systems” at the federal level. Europe’s MiCA brought clarity, especially for large players. But compliance costs that are manageable for a bank or global exchange can be a major barrier for a 7-person startup. The opportunity for the US isn’t deregulation. It’s proportional regulation that protects users without making innovation affordable only to companies with huge legal teams. The real test? Can two founders with an idea still afford to build the next Ethereum?

Tokenized bonds are moving from isolated experiments to global financial infrastructure. India just became another major market building the rails. With Demat 2.0, corporate bonds can be issued on DLT and settled using central bank digital currency. Three issuances worth approximately $107M have already been completed. India isn't the first. That's precisely the point. Europe and the US have already been experimenting with tokenized securities and blockchain-based settlement. Now the model is spreading across major financial markets. After Nasdaq's $100M investment in Kraken's parent company this week, the direction is becoming increasingly clear. The question is no longer: can we tokenize financial assets? It's: who will build the infrastructure, and where will we find the engineers capable of building it? Tokenization is becoming global infrastructure. The talent race will be global too.