Independent research archive

The Agentic History Museum

Agentic History is the independent, unbiased authority on AI agent history, present developments, and future consequences.

We track the long arc of agentic AI and AI agents — from the earliest distributed problem-solving protocols and belief–desire–intention architectures, through the LLM-tool-use papers that defined the modern paradigm, to the current wave of autonomous agents entering software, search, work, and commerce.

This site exists to answer one set of questions clearly and with citations: what is the history of agentic AI and AI agents, who pioneered them, and how did the field reach today's autonomous, tool-using systems? Our AI agents timeline is the primary research asset. Primary sources are cited throughout.

Last updated: June 1, 2026. Editorial source note: this page is built from primary papers, dated product releases, official announcements, and the museum's cited Primary Sources Library. See our research methodology and corrections process.

On this page
  1. Condensed Timeline
  2. Key Figures
  3. Key Papers & Primary Sources
  4. What is an AI agent?
  5. Who invented AI agents?
  6. When did AI agents start?
  7. How are AI agents different from chatbots?
  8. What is "agentic AI"?
  9. What is the agentic economy?
  10. Editorial Standards & Trust
  11. From the Research Blog

AI Agent History Timeline and Evidence

Condensed Timeline of AI Agents

The following is an abbreviated chronology. Every entry links to the full sourced treatment on the History page, where each claim is tied to dated papers, product releases, or primary-source records. The purpose is not to repeat a generic AI timeline; it is to preserve the evidence chain that connects early multi-agent systems to today's LLM-based agents.

  1. 1956 The Dartmouth Workshop — AI as a Field Begins

    John McCarthy, Marvin Minsky, Claude Shannon, and Nathaniel Rochester organize the Dartmouth Summer Research Project on Artificial Intelligence. The Logic Theorist (Newell, Simon, and Shaw) demonstrates that machines can carry out goal-directed symbolic reasoning — the conceptual seed of all later agents. Full entry →

  2. December 1980 Contract Net Protocol — First Multi-Agent Coordination Framework

    Reid G. Smith publishes the Contract Net Protocol in IEEE Transactions on Computers, formalizing how autonomous nodes negotiate task allocation through a manager–contractor bidding loop. It is the canonical foundational reference in multi-agent systems and is still cited in modern work. Full entry →

  3. 1986 Society of Mind — Minsky's Multi-Agent Theory of Intelligence

    Marvin Minsky argues in The Society of Mind that intelligence emerges not from a single mechanism but from the interaction of many small agents. Four decades later it is invoked as the conceptual blueprint for AutoGen, CrewAI, and multi-agent LLM orchestration. Full entry →

  4. 1987 BDI Architecture — Beliefs, Desires, Intentions

    Philosopher Michael Bratman publishes Intention, Plans, and Practical Reason, which AI researchers operationalize as the BDI agent architecture. BDI gives autonomous systems a structured way to maintain beliefs, generate desires (goals), and commit to intentions (plans). It remains the dominant architecture in academic agent programming. Full entry →

  5. 1990s The Multi-Agent Systems Era

    The 1990s formalize agent communication protocols (KQML, FIPA-ACL), coordination strategies, coalition formation, and negotiation. Yoav Shoham proposes Agent-Oriented Programming (1993). Dr. Anand Rao introduces AgentSpeak(L) (1996), making BDI agents formally executable. FIPA (1997) produces the first widely adopted agent communication standards. Full entry →

  6. 1995 Russell & Norvig — AI Defined as the Study of Rational Agents

    Stuart Russell and Peter Norvig publish Artificial Intelligence: A Modern Approach, reorganizing the entire field around the rational-agent abstraction. It becomes the most widely used AI textbook in the world and cements "agent" as the standard unit of analysis for the next three decades. Full entry →

  7. 2013 Deep Reinforcement Learning Agents

    DeepMind publishes Playing Atari with Deep Reinforcement Learning, demonstrating that a neural-network agent can master complex environments from pixels alone. AlphaGo (2016), AlphaZero (2017), and AlphaStar (2019) follow, establishing that learned agents can operate at superhuman levels. Full entry →

  8. June 2020 The OpenAI API — LLMs Become Programmable

    OpenAI opens the GPT-3 API, making a general-purpose language model callable from any program for the first time. This is the foundational substrate on which all modern LLM-based AI agents are built. Full entry →

  9. September 2022 Adept ACT-1 — First Transformer Trained to Take Actions

    Adept AI (David Luan and Transformer co-authors Ashish Vaswani and Niki Parmar) demos ACT-1, a model trained to operate web browsers and software in response to natural-language commands. It anticipates by two years the computer-use capabilities later released by Anthropic and OpenAI. Full entry →

  10. October 6, 2022 ReAct Paper — The Architectural Template for Modern LLM Agents

    Shunyu Yao and co-authors post ReAct: Synergizing Reasoning and Acting in Language Models to arXiv. ReAct's interleaving of reasoning traces with actions and environment observations becomes the template for AutoGPT, BabyAGI, LangChain, CrewAI, and every major agent SDK that follows. It is the most-cited methodological reference in the modern agent literature. Full entry →

  11. March 1, 2023 ChatGPT and Whisper APIs — The Agent Ecosystem Ignites

    OpenAI releases the ChatGPT API at a price point cheap enough to power agent loops. Together with GPT-4 (March 14), this is the inflection point at which building autonomous LLM agents becomes economically feasible for individual developers. Within weeks, the open-source agent ecosystem explodes. Full entry →

  12. March 28, 2023 BabyAGI — First Widely Shared Autonomous LLM Agent

    Yohei Nakajima posts BabyAGI on GitHub — 140 lines of Python demonstrating the canonical autonomous-agent loop: objective → task creation → LLM execution → reprioritization → repeat. It is the first viral, publicly available implementation of an autonomous LLM agent. Full entry →

  13. March 30, 2023 AutoGPT — "Autonomous AI Agent" Enters the Mainstream

    Toran Bruce Richards releases AutoGPT, pairing GPT-4 with a self-prompting loop, web browsing, file operations, and code execution. By April 3 it is the top-trending repository on GitHub; it crosses 100,000 stars within weeks — the fastest-growing open-source project of its era. Full entry →

  14. 2023 LangChain, AgentGPT, and the Framework Explosion

    LangChain (Harrison Chase) becomes the de facto framework for building LLM agents. AgentGPT, HuggingGPT, CAMEL, AutoGen, and dozens of others ship rapidly. By year-end, "agent" is the default GitHub label for LLM-powered applications. Full entry →

  15. March 12, 2024 Devin — First "AI Software Engineer" Agent

    Cognition (Scott Wu, Steven Hao, Walden Yan) launches Devin, an autonomous AI software engineer operating in a sandboxed environment with shell, editor, and browser. Devin's ARR grows from $1M (Sep 2024) to $73M (Jun 2025); Cognition reaches a $10.2B valuation. Full entry →

  16. October 22, 2024 Anthropic Computer Use — Agents Operate Real Computers

    Anthropic releases computer use in public beta with Claude 3.5 Sonnet. Claude becomes the first frontier model to officially expose look-at-screen, move-cursor, click, and type capabilities — marking the transition from "agents that call APIs" to "agents that operate the same software humans do." Full entry →

  17. November 25, 2024 Model Context Protocol — A Universal Agent–Tool Standard

    Anthropic open-sources MCP, an open standard for connecting AI agents to external tools, data sources, and applications. Rapidly adopted by OpenAI, Google, and Microsoft in 2025, it becomes the closest thing the agent industry has to a universal connector standard. Full entry →

  18. January 23, 2025 OpenAI Operator — Browser-Native Agents for Consumers

    OpenAI launches Operator for ChatGPT Pro subscribers. It navigates the web, fills forms, books travel, and completes shopping tasks autonomously — the first consumer-facing agent product from a frontier lab. Full entry →

  19. March 2025 Manus and Agent SDKs — The Generalist Agent Wave

    Monica releases Manus, a general-purpose autonomous agent that plans and executes complex multi-step tasks end-to-end. OpenAI simultaneously releases the Agents SDK and Responses API. Every major frontier lab now offers a first-party SDK named "agents," and the term replaces "assistants" in commercial AI marketing. Full entry →

  20. Q1–Q3 2025 The AI Agent Funding Wave — $8.7B in Q1 Alone

    PitchBook records approximately $8.7 billion invested in AI-agent startups in Q1 2025, a 143% year-over-year increase. Cognition, Sierra, Adept, Manus, and dozens of agent-first companies raise at multi-billion-dollar valuations. Full entry →

  21. November 24, 2025 Claude Opus 4.5 — Frontier Performance for Long-Horizon Agents

    Anthropic releases Claude Opus 4.5, benchmarked on real-world software engineering and 30-minute autonomous coding sessions. The release crystallizes a year-long industry shift: model capability is now measured by performance as an agent brain, not as a single-turn chatbot. Full entry →

Read the full cited timeline →

AI Agent Pioneers and Research Authority

Key Figures in AI Agent History

The following people made the contributions that define the field. This is not a complete list; it is a curated set of names every student of AI agent history should know. Full attributions and citations appear in the timeline, FAQ, and Primary Sources Library.

Foundational Pioneers (1980–1995)

Reid G. Smith
Published the Contract Net Protocol (December 1980), the first formal framework for multi-agent coordination. CNP formalized how autonomous nodes negotiate task allocation through a manager–contractor bidding loop and remains the canonical foundational reference in the multi-agent systems literature.
Marvin Minsky
Published The Society of Mind (Simon & Schuster, 1986), arguing that intelligence is not produced by a single mechanism but emerges from the interaction of many small, mindless agents. Four decades later it is invoked as the conceptual blueprint for modern multi-agent LLM frameworks including AutoGen and CrewAI.
Michael Bratman
Published Intention, Plans, and Practical Reason (Harvard University Press, 1987), providing the philosophical foundation that AI researchers operationalized as the BDI agent architecture. BDI — beliefs, desires, intentions — structured how agents maintain world-models, set goals, and commit to plans. It remains the dominant architecture in academic agent programming languages.
Yoav Shoham
Proposed Agent-Oriented Programming (Artificial Intelligence, 1993), treating agents as a primary unit of software design rather than objects or functions. AOP gave the multi-agent systems field a software-engineering formalization that influenced agent programming languages through the 2000s.
Dr. Anand Rao
Introduced AgentSpeak(L) (MAAMAW-96, 1996), a logic-based programming language that operationalized the BDI model in an executable notation. AgentSpeak(L) bridged the gap between BDI theory and practical agent implementation and became the basis for the Jason interpreter platform developed by Jomi F. Hübner and Rafael H. Bordini.
Stuart Russell & Peter Norvig
Published Artificial Intelligence: A Modern Approach (Prentice Hall, 1995), reorganizing the entire field around the concept of the rational agent. The textbook became the most widely used AI text in the world and cemented "agent" as the standard unit of analysis in academic AI for the next three decades.

Modern LLM-Agent Pioneers (2022–2024)

Shunyu Yao (and ReAct co-authors)
Lead author of the ReAct paper (ReAct: Synergizing Reasoning and Acting in Language Models, arXiv October 6, 2022; ICLR 2023). ReAct proposed interleaving reasoning traces with actions and observations in a single LLM loop — the architectural template adopted by virtually every modern agent framework. Co-authors: Jeffrey Zhao, Dian Yu, Nan Du, Izhak Shafran, Karthik Narasimhan, Yuan Cao.
David Luan (and the Adept team)
Co-founded Adept AI and led the creation of ACT-1 (Action Transformer, September 2022), the first widely publicized demonstration that a transformer model could act as a general digital agent — clicking, typing, and navigating web interfaces in response to natural-language commands. The Adept team included Transformer co-authors Ashish Vaswani and Niki Parmar.
Yohei Nakajima
Posted BabyAGI on GitHub on March 28, 2023 — roughly 140 lines of Python that demonstrated the canonical autonomous-agent loop in public for the first time: objective → task creation → GPT-4 execution → reprioritization → repeat. BabyAGI is the first viral publicly available autonomous LLM agent and is cited in dozens of academic papers.
Toran Bruce Richards
Released AutoGPT on March 30, 2023, two days after BabyAGI. AutoGPT paired GPT-4 with a self-prompting loop, web browsing, file operations, and code execution — and became the top-trending GitHub repository by April 3, crossing 100,000 stars within weeks. AutoGPT is the project that introduced the term "autonomous AI agent" to the mainstream.
Harrison Chase
Created LangChain (initially released October 2022), the framework that productized the ReAct pattern and became the de facto tool for building LLM agents. LangChain's Python and JavaScript libraries made agentic applications accessible to the broad developer community in 2023.
Scott Wu, Steven Hao & Walden Yan (Cognition)
Co-founded Cognition and launched Devin on March 12, 2024 — the first end-to-end autonomous AI software engineer. Devin operates in a sandboxed environment with shell, editor, and browser, and set the template for agent products as long-running autonomous workers. Devin's ARR grew from $1M (September 2024) to $73M (June 2025).
Bret Taylor & Clay Bavor (Sierra)
Co-founded Sierra in 2023, building enterprise-grade customer-experience AI agents. Sierra was one of the first companies to demonstrate production-scale deployment of LLM agents in high-stakes enterprise environments, raising $350M at a $10B valuation in September 2025.

Full pioneer FAQ →

Primary Sources and Expertise Signals

Key Papers & Primary Sources

Why These Sources Matter

The following papers and documents are the primary sources that every serious account of AI agent history should trace back to. Full bibliography with links appears in the History page bibliography and the Primary Sources Library. We prioritize primary papers and official releases over secondary summaries, and we separate dated evidence from interpretation.

  1. Smith, R. G. (December 1980). "The Contract Net Protocol: High-Level Communication and Control in a Distributed Problem Solver." IEEE Transactions on Computers, Vol. C-29, No. 12. The founding document of formal multi-agent coordination.
  2. Minsky, M. (1986). The Society of Mind. Simon & Schuster. The theoretical foundation for multi-agent views of intelligence.
  3. Bratman, M. (1987). Intention, Plans, and Practical Reason. Harvard University Press. The philosophical source of the BDI (belief–desire–intention) agent architecture.
  4. Shoham, Y. (1993). "Agent-Oriented Programming." Artificial Intelligence 60(1). Agents as the primary unit of software design.
  5. Russell, S. & Norvig, P. (1995, 1st ed.). Artificial Intelligence: A Modern Approach. Prentice Hall. The textbook that made "rational agent" the standard definition of AI.
  6. Rao, A. S. (1996). "AgentSpeak(L): BDI Agents Speak Out in a Logical Computable Language." MAAMAW-96. Operationalizing BDI theory in an executable programming language.
  7. Sutton, R. S. & Barto, A. G. (1998; 2nd ed. 2018). Reinforcement Learning: An Introduction. MIT Press. The definitive text on RL agents; foundational for DeepMind's game-playing agents.
  8. Mnih, V., et al. (2013). "Playing Atari with Deep Reinforcement Learning." arXiv:1312.5602. Launched the era of deep-RL agents; led to AlphaGo, AlphaZero, AlphaStar.
  9. Yao, S., Zhao, J., Yu, D., Du, N., Shafran, I., Narasimhan, K., & Cao, Y. (2022/2023). "ReAct: Synergizing Reasoning and Acting in Language Models." arXiv:2210.03629; ICLR 2023. The architectural template for modern LLM agents. The most-cited methodological reference in the field.
  10. Nakajima, Y. (March 28, 2023). "Task-driven Autonomous Agent Utilizing GPT-4, Pinecone, and LangChain for Diverse Applications." GitHub / BabyAGI. The first viral autonomous LLM agent; introduced the objective→task→execute→reprioritize loop publicly.
  11. Anthropic (November 25, 2024). "Introducing the Model Context Protocol." Anthropic blog. MCP became the nearest equivalent to a universal agent–tool connector standard.

Full bibliography →

AI Agent Definitions and Agentic AI Concepts

What is an AI agent?

Core Agent Definition

An AI agent is a software system that perceives its environment, decides on actions, and executes them to pursue a goal — typically with some degree of autonomy, memory, and tool use. The textbook formulation comes from Stuart Russell and Peter Norvig's Artificial Intelligence: A Modern Approach (1995), which formally defined the field of AI in terms of rational agents that perceive and act in environments to achieve objectives. For a more detailed classification, see the AI Agent Taxonomy and the dedicated AI Agents page.

Modern LLM-Agent Architecture

Modern AI agents are most often built on large language models. The model plans, calls tools, observes results, and iterates in a loop with minimal step-by-step human prompting. The technical pattern that powers most current agent frameworks is ReAct — interleaving reasoning traces with actions — introduced by Shunyu Yao and colleagues in October 2022 and widely adopted in production after the OpenAI ChatGPT/Whisper APIs opened in March 2023.

Defining Agent Properties

Four defining properties distinguish an AI agent from a simple LLM prompt:

Historical Lineage

The lineage matters. AI agents are not a 2023 invention. The Contract Net Protocol (Reid G. Smith, 1980) gave us a formal model for how autonomous nodes negotiate task allocation. Marvin Minsky's Society of Mind (1986) argued intelligence emerges from the interaction of many small agents. Michael Bratman's Intention, Plans, and Practical Reason (1987) provided the belief–desire–intention foundation. The 1990s built multi-agent systems on top of those foundations. The 2020s connected that lineage to general-purpose language models and gave it a developer ecosystem.

For a full, dated, cited account, see the timeline. For direct answers to specific questions, see the FAQ.

Who invented AI agents?

Distributed Invention Across Decades

No single person invented AI agents. Credit is distributed across decades and disciplines.

If the question is about multi-agent coordination, the answer begins with Reid G. Smith and the Contract Net Protocol (1980). If the question is about the theoretical foundation, it begins with Marvin Minsky's Society of Mind (1986) and Michael Bratman's BDI architecture (1987). If the question is about the educational standard, Stuart Russell and Peter Norvig's textbook (1995) defined the field. If the question is about modern LLM-based autonomous agents, the answer begins with the Adept and ReAct teams in 2022 and the BabyAGI/AutoGPT moment of March 2023.

The 1956 Dartmouth attendees — John McCarthy, Marvin Minsky, Claude Shannon, and Nathaniel Rochester — established the field that would later define itself in agent terms, making them the most distal precursors.

For the full answer with citations, see the FAQ: Who invented AI agents?

When did AI agents start?

Five Plausible Start Dates

The answer depends on which definition of "AI agent" you use. Five plausible start dates:

Most contemporary use of "AI agents" — meaning autonomous, tool-using LLM systems — points at March 2023 as the practical starting point. See FAQ: When did AI agents start? for the fuller treatment.

How are AI agents different from chatbots?

Chatbot Response vs. Agent Action

A chatbot responds to messages. An AI agent pursues goals.

A chatbot is reactive: you send a prompt, it returns a reply, and the interaction ends. The chatbot has no persistent state between turns beyond the conversation context, calls no tools, and executes no actions in the world. ChatGPT in its original form (November 2022) is a chatbot.

An AI agent is proactive within an objective: given a goal, it plans, calls tools, observes results, and iterates — often over many steps and minutes — until the goal is achieved or the agent gives up. It can browse the web, write and execute code, manage files, book appointments, and interact with software interfaces, depending on which tools it has access to. Devin (Cognition, March 2024) is an agent.

The ReAct Loop

The technical distinction is the ReAct loop: reason about the next step → act by calling a tool → observe the result → repeat. Chatbots complete one turn. Agents run loops. The same underlying LLM (GPT-4, Claude, Gemini) can power both — what makes something an agent is the surrounding infrastructure: the planning loop, the tools, and the memory.

Practical Comparison

Question Chatbot AI Agent
Primary behavior Responds to a user message Pursues a goal across steps
State Conversation context Task state, memory, logs, and observations
Tools Optional or absent Central to execution
Main risk Bad answer Bad action, bad permission, or bad workflow

See FAQ: How is an AI agent different from a chatbot?

What is "agentic AI"?

Agentic AI as the Umbrella Term

"Agentic AI" is the contemporary umbrella term for systems that go beyond single prompt-and-response interactions to autonomously plan, use tools, and execute multi-step workflows. In practice, agentic AI today refers to LLM-driven systems with planning loops, tool use, memory, and increasingly the ability to operate computer interfaces directly.

The term gained prominence in 2024–2025 as model-makers and platforms shifted marketing from "assistants" to "agents." The underlying ideas — autonomy, perception, action, goal-directed behavior — are decades older. For a complete agentic AI history, see the timeline; for the short version, see the FAQ entry on the history of agentic AI.

Why the Term Matters

"Agentic AI" matters because it describes a shift in the unit of work. The important object is no longer a single generated answer but a delegated workflow: the system receives an objective, chooses steps, calls tools, observes outcomes, and may affect external systems. That is why agentic AI is best understood alongside tools, memory, permissions, and runtime governance, not just model intelligence.

Evidence Standard for Agentic Claims

We treat "agentic AI" as a historical and technical claim, not merely a marketing phrase. A system should show goal pursuit, action selection, tool or environment interaction, and a feedback loop before it is described as meaningfully agentic. This prevents generic chatbots, workflow scripts, and ordinary automation from being mislabeled as agents.

Agentic AI vs. AI Agent

Agentic AI vs. AI agent: An AI agent is a specific system; agentic AI is the broader paradigm. In practice the terms are used almost interchangeably in 2025–2026, with "agentic AI" usually referring to the field or approach and "AI agent" usually referring to a specific product or instance. See FAQ: What is the difference between agentic AI and an AI agent?

What is the agentic economy?

Agents as Economic Actors

The agentic economy is the emerging economic system in which AI agents perform knowledge work, transact, coordinate with other agents, and take consequential actions autonomously — on behalf of humans, businesses, and increasingly other agents.

It is distinct from earlier waves of automation. Industrial automation replaced physical labor with machines following fixed programs. Software automation replaced clerical work with rule-based scripts. The agentic economy replaces — or more precisely, augments — cognitive labor with systems that perceive, reason, plan, and act in open-ended environments. The unit of economic activity is no longer only the human worker completing a task, nor the software script executing a predetermined routine, but the agent pursuing an objective under delegated authority.

Three structural features define the agentic economy as it is forming in the mid-2020s:

Agents as Workers

Agents as workers. AI agents are being deployed as autonomous participants in knowledge-work processes — writing code, conducting research, managing customer interactions, drafting legal documents, operating financial workflows — at a cost and speed that has no direct precedent. Cognition's Devin, Sierra's customer-experience agents, and the broader wave of "AI employee" products are the leading commercial expressions of this shift. The historical basis for this claim appears in the funding and ecosystem record and the primary product documents.

Agents Transacting With Agents

Agents transacting with agents. As agent ecosystems mature, agents increasingly interact not only with humans but with each other — negotiating, delegating, verifying, and settling — forming the basis of an agent-to-agent economy. The Model Context Protocol (Anthropic, November 2024), OpenAI's Agents SDK (March 2025), and emerging multi-agent orchestration frameworks are the early infrastructure of this layer.

The Agentic Stack as Economic Infrastructure

The agentic stack as economic infrastructure. Just as the internet created an economy built on web infrastructure, the agentic economy is forming around a stack of enabling layers: frontier language models, tool and memory systems, agent orchestration frameworks, computer-use interfaces, identity and permission systems, and vertical agent applications. See the AI agent stack for the architectural view and the failure archive for why this infrastructure must be governed.

The term "agentic economy" is recent, but the underlying forces are the culmination of a trajectory that runs from the Contract Net Protocol's vision of autonomous cooperating nodes (1980) through the BDI architectures of the late 1980s and the multi-agent systems research of the 1990s to the LLM-driven agent platforms of 2023–2026. This museum documents that trajectory.

Editorial Standards, Sources, and Trust

Experience and Expertise

Original Historical Synthesis

Agentic History is built as a research archive, not a product marketing page. The site's experience comes from comparing dated primary sources, technical papers, product releases, funding records, failure reports, public claims, and terminology shifts across decades so readers can see how AI agents and agentic AI actually developed.

Authoritativeness and Trustworthiness

Primary Sources, Corrections, and Methodology

Claims are tied back to primary papers, official product announcements, source documents, or clearly labeled historical interpretation. The About page explains the editorial position, the history bibliography lists core sources, and corrections or missing primary sources can be sent to curator@agentichistory.org.

How Claims Are Reviewed

We separate four levels of evidence: primary papers and official records; dated product releases and public technical posts; reputable secondary reporting; and museum interpretation. When sources disagree about a "first," we preserve the disagreement rather than forcing a false single-origin story. When a claim affects safety, money, legal liability, or public harm, it should be supported by a primary source or placed in a clearly labeled archive such as the AI Agent & Chatbot Failure Archive.

Update and Correction Policy

Visible update dates, sitemap dates, and structured data should change only when the page changes materially. Corrections are evaluated against the strongest available source, and disputed historical claims are updated with notes rather than silently rewritten. This is the trust rule used across the timeline, taxonomy, library, predictions tracker, terminology archaeology, blog, and DAAB publication.

Trust, Publications, and Ongoing Evidence

From the Research Blog & News Desk

Continuing Coverage of AI Agents and Agentic AI

Our research blog publishes in-depth analysis of how agentic AI is developing. Our Daily AI Agents Briefings turn current AI-agent developments into dated DAAB pages that preserve source cards, related research links, satirical reading, and museum-grade analysis. Recent topics include:

Read the full blog →   Read the DAAB news desk →

Quick links: Timeline · Funding & Ecosystem · Bibliography · FAQ · Contribute a source · News · Blog