2025: A Defining Year for Web3 + AI
We welcome the new year with a retrospective of the most significant Web3 + AI moments of 2025.
Hello, everyone, and welcome to all newly-joined subscribers!
This is The Web3 + AI Newsletter, your definitive guide to the world of Decentralized AI (DeAI/dAI)!
Happy New Year 🥳 I’m grateful to have welcomed another new year in your company. May 2026 bring you peace, joy, prosperity, and a bunch of exciting adventures!
I’m kicking off 2026 by recapping last year’s Web3 + AI moments, whose implications will continue to greatly impact the entire sector.
Thank you for being here! Let’s dive in.
Key Web3 + AI Events of 2025
Ethereum Foundation’s dAI Team and “Trustless Agents”
In October 2025, the Ethereum Foundation doubled down on decentralized AI by launching a dedicated dAI team, led by Davide Crapis and aimed at making Ethereum the primary settlement and coordination layer for the emerging AI agent economy. Central to this effort is ERC-8004, a “Trustless Agents” standard that enables AI agents to discover, verify, build reputation, and transact with one another without pre-existing trust.
Developed as a cross-industry collaboration involving Ethereum, Google, MetaMask, and others, ERC-8004 complements Google’s Agent-to-Agent (A2A) and Agent Payments (A2P) protocols while addressing their trust limitations. The initiative positions Ethereum as uniquely suited to power agentic commerce by providing native payment rails, identity, and trust infrastructure - capabilities missing from existing AI frameworks.
Coinbase’s x402 Payment Protocol
2025 was a spectacular year for agentic payments, with Coinbase introducing x402, a new open protocol that makes stablecoin micropayments a native part of web interactions.
This lets apps, services, and AI agents automatically execute on-chain payments for access to APIs, data, or services without accounts, subscriptions, or middlemen, unlocking true machine-to-machine commerce and pay-per-use monetization.
The initiative quickly gained industry support from partners like Circle, NEAR Protocol, Anthropic, Amazon Web Services (AWS), and Cloudflare , which helped establish the x402 Foundation to govern and advance the standard.
Adoption surged as the protocol expanded beyond Coinbase’s Base Chain to multiple EVM networks and Solana, driving rapid transaction growth and ecosystem interest. However, challenges remain before it can fully power a mature machine economy, and further infrastructure (like trust, discovery, and quality services) is still developing.
Bittensor’s dTAO Upgrade and Halving
In 2025, the open-source AI network Bittensor / Opentensor Foundation performed not one, but two pivotal economic overhauls. First, the Dynamic TAO upgrade was enacted, followed by the first-ever $TAO halving event.
dTAO, or dynamic TAO, marks a major change in Bittensor’s tokenomics by giving each of the network’s specialized subnets its own token, paired with TAO in a dedicated liquidity pool.
Instead of subnets competing for a shared TAO reward and relying on a single group of validators to rank them, each subnet’s value now rises or falls based on genuine demand via the users, stakers, and validators who adopt it.
In mid-December, Bittensor’s first token halving reduced daily $TAO emissions from 7,200 to 3,600, and aligned the protocol with Bitcoin-style scarcity mechanics.
The halving shook up the network’s subnet ecosystem: with fewer emissions available, capital and rewards now flow toward higher-performing subnets, while weaker or “zombie” subnets struggle for liquidity or become inactive.
Additionally, mining and validator rewards tightened, potentially leading to consolidation among the most efficient participants. Supporters argue this will strengthen the network’s economic health over time, even if short-term volatility persists.
Crypto Treasuries On The Rise
In 2025, digital asset treasuries (DATs), where public companies hold cryptocurrencies as core balance-sheet assets, shifted from niche experiments to a mainstream corporate finance trend. Major firms expanded beyond Bitcoin to include Ethereum and even dAI coins in their treasuries, generating billions in net inflows and signaling widespread institutional adoption.
Strategy (formerly MicroStrategy) remained the largest holder, with over $62B in BTC, while new players like BitMine Immersion Technologies, Sharplink, Ether Machine (ETHM), and Forward Industries (Nasdaq: FWDI) amassed significant crypto holdings. Meanwhile, in the dAI vertical, TAO Synergies Inc. and Synaptogenix bet on Bittensor’s $TAO token, and Predictive Oncology launched a $344M Aethir DAT.
This evolution was supported by clearer regulations, easier capital access, and confidence in digital assets as stores of value. However, heavy concentration of DATs among a few firms and associated risks, such as volatility and investor scrutiny, highlight ongoing challenges. The next stage for DATs is to generate yield and integrate digital assets into real operational use cases like staking, settlement, and liquidity management, rather than merely stockpiling tokens.
BTC Mining vs. AI / HPC Data Centers
Over the past year, I’ve covered the same piece of news over and over again - the only thing that changed was the protagonists. With BTC mining rewards declining and demand for AI computing resources growing incessantly, miners were gradually shifting to AI and HPC (high-performance computing) operations, seeking higher profits and less uncertainty.
What’s more, those who made the switch performed better financially: IREN emerged as the top performer with roughly +300% year-to-date gains, driven by major GPU cloud deals and support from Microsoft. Other AI-focused miners like Cipher Digital and Hut 8 also posted strong returns, buoyed by AI hosting partnerships and large data-center projects.
By contrast, traditional Bitcoin-heavy miners lagged: Bitdeer (NASDAQ: BTDR) saw its stock fall about 50%, affected by a larger-than-expected Q3 loss and delays in ASIC chip development, while peers like Marathon Digital experienced declines without significant AI diversification. The trend highlights how miners leveraging AI/HPC demand have outperformed pure-play crypto mining operations in 2025.
Now, with its 2026 energy forecast, BlackRock warns of a looming energy clash between AI data centers and Bitcoin miners. The company highlights that AI-driven infrastructure could consume a very large share of U.S. electricity in the coming years, potentially reshaping how power is allocated across industries.
Bitcoin mining has traditionally competed for cheap, flexible energy by absorbing surplus generation and curtailing during peak demand, but AI workloads require constant, reliable baseload power, which could squeeze miners’ access to favorable grid capacity and elevate energy costs. As grid constraints tighten, miners may have to adapt by integrating more closely with utility planning or pivoting toward AI/HPC hosting services, changing the competitive landscape for energy-intensive tech sectors.
The question remains: what will happen with Bitcoin when all of its major miners pivot to AI?
Thank you for reading! I hope you found this article insightful.
If you’d like to keep up with the latest developments at the burgeoning intersection of blockchain and AI, I invite you to subscribe.
And if you find value in my writing, consider supporting the newsletter. I’d greatly appreciate it.
The Web3 + AI Book Club is live on Fable. Follow the link below and come join our literary community.
I'm looking forward to connecting with fellow Crypto x AI enthusiasts, so don't hesitate to share your thoughts here or reach out on LinkedIn, Instagram, or Firefly.social.
Disclaimer: The content of this newsletter is provided for informational and educational purposes only. Nothing contained herein should be construed as financial advice or as a recommendation to buy, sell, or hold any of the companies or assets mentioned.
Please don’t take my views at face value. Instead, do your own research (DYOR), think critically, and share your perspective so we can challenge ideas, learn from one another, and arrive at better conclusions through thoughtful discussion.