Tootfinder

OpenAI updates its coding agent Codex with internet access, turned off by default, and expands availability to ChatGPT Plus users (Simon Willison/Simon Willison's Weblog)
https://simonwillison.net/2025/Jun/3/codex-agent-internet-access/

Codex agent internet access
Sam Altman, [just now](https://twitter.com/sama/status/1930006856019390521): > codex gets access to the internet today! it is off by default and there are complex tradeoffs; people should read about the risks carefully and …

#Gartner prognostiziert: Bis 2027 werden 40% der KI-Agenten-Projekte wieder eingestampft. Der Hype um autonome #KI #Agenten entpuppt sich als teurer Irrweg – zu hohe Kosten, unklare Vorteile, mangelnde Reife…

Künstliche Intelligenz: 40 Prozent aller KI-Agenten bis 2027 wieder abgeschafft - Golem.de
Viele KI-Agenten gehen nicht über die Funktionsweise von skriptbasierten Bots und virtuellen Assistenten hinaus.

The Automated but Risky Game: Modeling Agent-to-Agent Negotiations and Transactions in Consumer Markets
Shenzhe Zhu, Jiao Sun, Yi Nian, Tobin South, Alex Pentland, Jiaxin Pei
https://arxiv.org/abs/2506.00073

The Automated but Risky Game: Modeling Agent-to-Agent Negotiations and Transactions in Consumer Markets
AI agents are increasingly used in consumer-facing applications to assist with tasks such as product search, negotiation, and transaction execution. In this paper, we explore a future scenario where both consumers and merchants authorize AI agents to fully automate negotiations and transactions. We aim to answer two key questions: (1) Do different LLM agents vary in their ability to secure favorable deals for users? (2) What risks arise from fully automating deal-making with AI agents in consum…

A post from the archive 📫:
Validate your identity on Bluesky with your GitHub account
https://www.poppastring.com/blog/validate-your-identity-on-bluesky-with-your-github-account

If an "undercover" officer is seen and known to the press then I would not hesitate to publish their face image and identity - in part because they have failed actually be "undercover".
That said, the fact that agents can't be identified as actual Federal agents, much less identified down to the individual actor when they do ill, is simply unacceptable.
TSA at least has numbers on their "name" badges, so one can make a complaint about a person …

The ICE agents disappearing your neighbors would like a little privacy, please
After taking four people into custody at SF’s immigration court, ICE wanted The Standard to conceal agents’ faces.

A Study on the MCP x A2A Framework for Enhancing Interoperability of LLM-based Autonomous Agents
Cheonsu Jeong
https://arxiv.org/abs/2506.01804 https://

A Study on the MCP x A2A Framework for Enhancing Interoperability of LLM-based Autonomous Agents
This paper provides an in-depth technical analysis and implementation methodology of the open-source Agent-to-Agent (A2A) protocol developed by Google and the Model Context Protocol (MCP) introduced by Anthropic. While the evolution of LLM-based autonomous agents is rapidly accelerating, efficient interactions among these agents and their integration with external systems remain significant challenges. In modern AI systems, collaboration between autonomous agents and integration with external t…

Google is ending PayPal account integration in Google Wallet for US users, effective June 13, after stopping new PayPal account linking in April (Mishaal Rahman/Android Authority)
https://www.androidauthority.com/paypal-killing-google-wallet-inte…

PayPal is killing Google Wallet integration for users in the US
PayPal has long let you link your PayPal account to Google Wallet, but that integration is ending soon in the US. Here's what'll happen.

This https://arxiv.org/abs/2410.04444 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csAI_…

Gödel Agent: A Self-Referential Agent Framework for Recursive Self-Improvement
The rapid advancement of large language models (LLMs) has significantly enhanced the capabilities of AI-driven agents across various tasks. However, existing agentic systems, whether based on fixed pipeline algorithms or pre-defined meta-learning frameworks, cannot search the whole agent design space due to the restriction of human-designed components, and thus might miss the globally optimal agent design. In this paper, we introduce Gödel Agent, a self-evolving framework inspired by the Göde…

AgentAuditor: Human-Level Safety and Security Evaluation for LLM Agents
Hanjun Luo, Shenyu Dai, Chiming Ni, Xinfeng Li, Guibin Zhang, Kun Wang, Tongliang Liu, Hanan Salam
https://arxiv.org/abs/2506.00641

AgentAuditor: Human-Level Safety and Security Evaluation for LLM Agents
Despite the rapid advancement of LLM-based agents, the reliable evaluation of their safety and security remains a significant challenge. Existing rule-based or LLM-based evaluators often miss dangers in agents' step-by-step actions, overlook subtle meanings, fail to see how small issues compound, and get confused by unclear safety or security rules. To overcome this evaluation crisis, we introduce \sys, a universal, training-free, memory-augmented reasoning framework that empowers LLM evaluator…

The Ultimate Test of Superintelligent AI Agents: Can an AI Balance Care and Control in Asymmetric Relationships?
Djallel Bouneffouf, Matthew Riemer, Kush Varshney
https://arxiv.org/abs/2506.01813

The Ultimate Test of Superintelligent AI Agents: Can an AI Balance Care and Control in Asymmetric Relationships?
This paper introduces the Shepherd Test, a new conceptual test for assessing the moral and relational dimensions of superintelligent artificial agents. The test is inspired by human interactions with animals, where ethical considerations about care, manipulation, and consumption arise in contexts of asymmetric power and self-preservation. We argue that AI crosses an important, and potentially dangerous, threshold of intelligence when it exhibits the ability to manipulate, nurture, and instrumen…