AI Agents Need Data Ecosystems, Not Marketplaces (2 of 2) (Tech Strategy)

Posted on by

In Part 1, I raised the question of whether AI Agents need marketplaces and other platform business models to enable interactions. Or can they operate with different business models.

My final point was that while AI Agents are definitely going to be active on marketplaces, they can also likely interact (and transact) with data ecosystems.

An Introduction Data Ecosystem Orchestration Platforms

Computers create networks based on standardized protocols (such as TCP). We often call these protocol networks. As opposed to social networks, made of connected humans.

Robert Metcalfe (the co-founder of the Ethernet and 3Com) famously said that the power of a protocol network increases exponentially with the number of connected computers. Each additional node creates many more connections.

Compare such protocol networks to marketplaces or other digital platforms? We see lots of protocol networks for devices to interact. But do we see marketplaces or other digital platforms for computers to interact?

Sort of.

In web3, we do see platform business models built on top of protocol networks (i.e., blockchains).

But mostly no.

So what do AI agents need to interact and transact with each other?

Do they need platforms? Protocol networks? Both?

Think about what we need for AI agents to interact. We need:

  1. AI-agent interoperability: Something to allow different AI agents to interact and collaborate.
  2. Tokenized data: You need to create standardized, tradable, and secure data assets.
  3. Data provenance: You need to ensure the integrity, quality, and origin of data.
  4. Transparent and explainable AI decision-making: This helps to build trust in AI-driven outcomes.
  5. Community-driven data curation: Something to encourage data sharing, validation, and improvement.

This doesn’t really sound like a platform business model as discussed. It sounds more like a data ecosystem, plus some tools and services.

And we also need more parties than just the AI agents. We need:

  1. AI Agents: They are consuming data, making decisions, doing interactions, doing transactions and generating insights. Such as buyers and sellers.
  2. Data Providers: They supply the fuel for AI decision-making.
  3. Data Processors: Who handle data storage, processing, and analytics
  4. Human Users: Who benefit from AI-driven services and provide feedback. This may not be necessary in lots of cases.

We really need all of these groups to interact in a data ecosystem for interactions to happen.

Enter the idea of Data Ecosystem Orchestration (DEO) platforms. These are data ecosystems that enable the coordination of these multiple parties. And fulfill the requirements for interoperability, tokenized data, data provenance and data curation.

If this is how marketplaces look…

CloudKitchens

Here’s how I think about DEOs:

Photo by Alina Grubnyak on Unsplash

Going forward, I fully expect AI agents will operate within platforms, as buyers and sellers.

But I also expect them to be using DEOs, where they contribute to the creation of a decentralized, self-sustaining data ecosystem. Where value is generated through the interactions between humans, AI agents, and data. I’ll provide some examples below.

One last point on DEOS. I don’t expect them to be static. We want such data ecosystems to grow and improve. They are evolving ecosystems. So, we also need to think about which DEOs would be more supportive of:

  • Data-driven innovation: Does it accelerate the development of new AI applications and services?
  • AI-driven entrepreneurship: Does it enable the creation of new businesses and revenue streams?
  • Improved decision-making: Does it enhance the accuracy, efficiency, and transparency of AI-driven decisions and interactions?

Examples: Marketplaces to DEOs

We’re still in the early days of AI agents interacting with each other. And of buying and selling products / services. Most of the initiatives today are about enabling AI agents to interact with existing e-commerce platforms, marketplaces, and other online platforms. Here are a few examples:

1. E-commerce Platforms with AI Integration.

  • Amazon Alexa. A consumer can tell Alexa to search and buy things for you on Amazon. You are basically giving a voice command to an AI agent which then acts as your personal shopper. Not so different than a search engine.
  • Google Assistant: This is similar. But it integrates with more e-commerce platforms, including Google Express. It enables AI agents to purchase products on behalf of users.
  • Facebook Marketplace: Which allows AI agents to interact with the platform, enabling users to buy and sell products using chatbots and other AI-powered tools.

2. AI-Powered Marketplaces. 

  • This a more complete integration. Alibaba and JD are already providing AI Agents to merchants on their platforms. And Alibaba is starting to enable DingTalk to operate as an AI personal assistant.
  • Shopify is another example. They offer an AI-powered marketplace that enables merchants to create custom chatbots and AI agents to interact with customers and facilitate transactions.
  • Rakuten: A Japanese e-commerce platform that uses AI agents to facilitate transactions and provide personalized recommendations to users.
  1.  
  1.  

This is pretty much how I think about AI-Powered Marketplaces.

3. Decentralized Marketplaces. 

  • This is more theoretical. It feels a lot like web3 business models (also theoretical).
  • OpenBazaar: This is a decentralized marketplace that enables AI agents to buy and sell products using blockchain technology and cryptocurrency.
  • District0x: a decentralized marketplace that uses AI agents to facilitate transactions and provide personalized recommendations to users.
  1.  

4. Virtual Assistant Marketplaces. 

  • This is about finding AI Agents to use.
  • Google Actions: a platform that enables developers to create AI-powered virtual assistants that can interact with users and facilitate transactions on behalf of e-commerce platforms.
  • Amazon Alexa Skills: a platform that enables developers to create AI-powered virtual assistants that can interact with users and facilitate transactions on behalf of e-commerce platforms.
  • Alibaba is launching their marketplace for AI agents, with DingTalk AI being one of the offerings.

5. AI Agent-Specific Marketplaces. 

  • This is interesting to think about. An AI powered marketplace is for humans and AI agents to interact. This is only for AI Agents. But mostly for models, data and compute resources.
  • ai: a decentralized marketplace that enables AI agents to buy and sell products and services, including data, compute resources, and machine learning models.
  • SingularityNET: a decentralized marketplace that enables AI agents to buy and sell AI services, including machine learning models, data, and compute resources.

Here’s how I think about this.

These marketplaces and platforms are still in the early stages of development, and the landscape is rapidly evolving. As AI agents become more sophisticated and ubiquitous, we can expect to see more marketplaces and DEO platforms emerge that cater to their needs and capabilities.

But the interesting question is whether existing marketplaces will be replaced or complemented with entirely new ways of interacting such as data ecosystems and DEOs.

Examples of Data Ecosystems

While the concept of Data Ecosystem Orchestration (DEO) platforms is still emerging, there are existing platforms and initiatives that share similarities with the DEO vision. These platforms can be seen as precursors or building blocks for the development of more comprehensive DEO platforms.

1. Data Marketplaces

  • DataBroker: A decentralized data marketplace for buying and selling data.
  • Ocean Protocol**: A platform for sharing and monetizing data, using blockchain and tokenization.
  • IOTA Data Marketplace**: A platform for sharing and selling data, using distributed ledger technology.

2. Data Exchange Platforms:

  • Dawex: A platform for data exchange, monetization, and collaboration.
  • Data Republic: A platform for secure, governed data exchange and collaboration.

3. Blockchain-based Data Platforms:

  • Filecoin: A decentralized storage network, using blockchain and tokenization.
  • InterPlanetary File System (IPFS)**: A decentralized storage and sharing platform.

4. Data Governance and Orchestration Platforms:

  • Collibra: A platform for data governance, quality, and lineage.
  • Informatica: A platform for data integration, governance, and management.

While these platforms address specific aspects of the DEO vision, they do not yet provide a comprehensive, integrated solution for orchestrating data ecosystems. The development of DEO platforms will likely involve the convergence of these existing platforms, as well as the creation of new technologies and innovations.

DEO and Data Ecosystem Challenges

Ok. That’s where my thinking is right now. Very fuzzy I know. But I think I’m asking the right questions.

A final question here is what are the big challenges in the development of AI agent marketplaces and DEOs. Here’s a list of some to think about.

  • Trust and Security: It is important to establish trust and security protocols so AI agents can interact with marketplaces and e-commerce platforms securely and reliably.
  • Standardization: It’s important to establish common standards for AI agent interactions with marketplaces and e-commerce platforms. This creates a collective action problem.
  • Interoperability: Enabling seamless interaction between different AI agents and marketplaces is important.
  • Regulatory Compliance: AI agent marketplaces need to comply with regulations and laws, such as consumer protection and data privacy laws.

***

That’s where my thinking is. We’ll see what emerges.

Cheers, Jeff

——-

Related articles:

From the Concept Library, concepts for this article are:

  • AI Agents
  • Marketplace Platform
  • Data Ecosystems
  • Protocol Networks

From the Company Library, companies for this article are:

  • N/a

——–

I write, speak and consult about how to win (and not lose) in digital strategy and transformation.

I am the founder of TechMoat Consulting, a boutique consulting firm that helps retailers, brands, and technology companies exploit digital change to grow faster, innovate better and build digital moats. Get in touch here.

My book series Moats and Marathons is one-of-a-kind framework for building and measuring competitive advantages in digital businesses.

This content (articles, podcasts, website info) is not investment, legal or tax advice. The information and opinions from me and any guests may be incorrect. The numbers and information may be wrong. The views expressed may no longer be relevant or accurate. This is not investment advice. Investing is risky. Do your own research.

Leave a Reply