+ Main Stage
Main Stage / AI x Blockchain Day
How to Prepare for the Future of Work
Agents, workflows, marketplaces, and the AI x Blockchain bridge
AI x Blockchain Day / Mexico City
+ The thesis
The future of work is not just people using AI tools.
It is work becoming programmable.
People will not only operate software.
People will design, supervise, and participate in agentic workflows.
+ The shift
The interface is moving from buttons to intent.
Before
- Open the app
- Move the data
- Click the buttons
- Export the file
- Send the update
Now
- Define the goal
- Provide context
- Set constraints
- Review the output
- Approve the risk
Language became an interface for work.
+ Executable work
If software can understand work, work can be packaged into repeatable systems.
A task becomes a process.
A process becomes a workflow.
A workflow becomes an executable capability.
+ From chatbot to agent
Not every AI system is an agent.
Chatbot
Answers questions.
Copilot
Assists a human while they work.
Agent
Takes a goal, uses tools, checks state, and moves work forward within constraints.
+ New human role
Agents execute more. The human moves up the stack.
The agent can
- Understand a task
- Break it into steps
- Use tools and APIs
- Read and write files
- Produce artifacts
- Ask for approval when risk is high
The person moves to
- Defining goals
- Setting standards
- Reviewing outputs
- Approving risk
- Improving the system
- Directing systems of work
+ Example
One goal. Many steps.
"Prepare a client proposal" can become an agentic workflow.
- 01 Gather client context
- 02 Research the company
- 03 Identify the problem
- 04 Draft scope
- 05 Estimate effort
- 06 Prepare commercial document
- 07 Route for human approval
- 08 Send only after approval
+ Workflows
Workflows make agents useful.
The value is not one AI answer. It is a repeatable system that produces better work over time.
- 01 Reduce ambiguity
- 02 Define inputs and outputs
- 03 Measure quality
- 04 Add human checkpoints
- 05 Coordinate handoffs
- 06 Repeat what works
- 07 Improve the system over time
+ Where this shows up
Any repeated work can become a candidate.
- 01 Sales qualification
- 02 Support triage
- 03 Research memos
- 04 Code review
- 05 Legal intake
- 06 Invoice reconciliation
- 07 Hiring pipelines
- 08 Community operations
- 09 Onchain transaction monitoring
+ The new professional skill
The new skill is designing work for agents.
- 01 Decompose tasks
- 02 Specify outputs
- 03 Choose tools
- 04 Create review gates
- 05 Evaluate AI work
- 06 Improve workflows
- 07 Package capabilities
+ From jobs to capabilities
Work is becoming modular.
Job
A broad role.
Task
A specific unit of work.
Workflow
A repeatable process.
Agent service
An executable capability.
+ Marketplaces
Marketplaces sell outcomes.
They may sell
- Human expertise
- AI agents
- Workflow templates
- Data, APIs, and tools
- Verification and reputation
People buy
- Qualified leads
- Researched reports
- Reviewed contracts
- Automated support flows
- Operational leverage
People do not buy "AI". They buy reliable results.
+ Trust
If agents do work across marketplaces, we need trust infrastructure.
- 01 Who performed the work?
- 02 What data was used?
- 03 What changed?
- 04 Was the output approved?
- 05 Who gets paid?
- 06 What reputation follows the work?
+ AI x Blockchain
AI gives capability. Blockchain gives guarantees.
AI gives capability
- Generate
- Decide
- Act
- Automate
Blockchain gives guarantees
- Verify
- Record
- Own
- Pay
- Coordinate
+ The four bridges
Identity, provenance, payments, and coordination.
Identity
Which human, company, or agent performed the action?
Provenance
Where did the data or output come from?
Payments
How does an agent charge or receive payment?
Coordination
How do multiple parties trust shared state?
+ Agentic commerce
Example: AI research agent
- 01 A client requests a report
- 02 The agent gathers sources
- 03 The agent produces the report
- 04 Provenance is attached
- 05 Human approval is recorded
- 06 Payment is settled through a wallet or contract
- 07 Reputation follows the completed work
- 08 The workflow can be reused or sold
+ How to prepare
Do not guess every future job title. Build capabilities that survive the transition.
Skill stack
- AI literacy
- Workflow design
- Tool and API thinking
- Data judgment
- Blockchain fundamentals
- Service packaging
What to practice now
- Pick one repeated process
- Turn it into a checklist
- Turn the checklist into a workflow
- Connect AI and a review gate
- Add trust where it matters
+ Closing thesis
Main Stage / Mexico City
The people who win are not the ones who only use AI.
They are the ones who learn to design work for AI, verify it with trustworthy systems, and package it into useful services.