SKALE School Episode 6: Virtuals - Recap

This week’s SKALE School dug into a simple question with big consequences: if agents are going to do real work onchain, such book flights, move funds, check facts, run workflows, what has to be true about the products we build and the infrastructure they run on? The episode featured host Ben Davey with Chris Johnson from Virtuals Protocol and Sawyer Cutler, SKALE’s Head of Developer Success.

In This Episode of SKALE School 

SKALE School is a recurring, conversational livestream where builders and product leaders unpack one topic deeply. This episode focused on autonomous agents, practical adoption, and SKALE’s Expand on Base architecture. Some high level takeaways from the episode include:

1) Meet users where they are

A through line from the conversation: mainstream users will not adopt crypto UX just to use an agent. They care that something gets done cheaply and reliably, not which rails it used under the hood. The panel’s view was clear: the average person does not care that it is crypto, they care about delegation and utility. Build for outcomes and hide the plumbing.

Why it matters: treating distribution as a first class problem reshapes product choices. Simple, repeatable agent tasks beat ornate everything agents, and the infrastructure should be invisible.

2) Swarms and SLMs beat one big model on cost and accuracy

The panel argued for small, specialized models coordinated in swarms. Many tasks such as fact checking need high accuracy at tiny unit costs, often fractions of a cent. That becomes critical when you scale from five checks a day to five hundred. Specialization and orchestration keep both accuracy and budgets in line.

Why it matters: this is not just an architecture preference. It is an economic requirement if agents are calling agents all day.

3) Per call economics beat subscriptions if base costs compress

The team sees a world where we pay per use or per call for almost everything and still beat subscription economics. That only works if the base layer delivers massive throughput at near zero cost, and can elastically scale when a swarm spikes to a huge workload and then scale back down.

Why it matters: if infrastructure cannot compress cost and expand supply on demand, per call business models break and so does the promise of agentic automation.

4) SKALE Expand on Base unlocks elastic blockspace and zero gas usage

SKALE explained Expand this way: SKALE Manager is a set of smart contracts that orchestrate SKALE chains and can be deployed on Base or any EVM. That lets builders spin up many SKALE chains inside the Base ecosystem, inheriting assets and tooling while gaining zero gas usage and instant finality. On testnet, SKALE has deployed x402 capable versions of USDT, WBTC, WETH, and SKL as bridged Base assets, along with trustless agent contracts so agents can be registered and paid on a zero gas chain.

Why it matters: builders get the distribution and liquidity of Base paired with SKALE’s fee less, high throughput execution. That is exactly the combo per call and swarm patterns demand.

5) Agents as customers and owners of blockspace

Because SKALE Manager is just smart contracts, agents can buy their own chain. A virtual agent could provision a SKALE chain in minutes, run high volume workflows for a weekend, bridge assets in and out, and then spin the chain down. In that framing, blockchains are built for machines. Humans benefit from what those machines accomplish.

Why it matters: this flips the customer model. Agents become primary consumers of blockspace, optimizing for cost and latency like a savvy ops team would.

Key Takeaway for Builders

Two pragmatic takeaways for teams building today:

• Reduce agent setup and iteration steps. If spinning up and improving an agent takes a dozen steps, most ideas stall at cool demo. Lower the steps, lean on shared discovery layers, and design for agents using agents.
  
• Let costs drive architecture. When each action must cost less than a ten thousandth of a cent, swarms plus SLMs win, and infra choices such as zero gas chains and elastic blockspace are existential.

Conclusion

The agent economy is getting practical. The episode’s through lines were consistent. Meet users where they are. Favor swarms and small models to keep quality and unit costs in check. And use infrastructure that treats machines as first class citizens. SKALE’s Expand on Base shows what that looks like in practice with zero gas usage, instant finality, x402 ready assets, and the bold idea that agents can be direct customers of blockspace. If you are building, ship small, specialize, and let the rails do their job so your users never have to think about them.