How Generative AI Is flooding the web with deepfakes and disinformation | TechCrunch
As generative AI tools become more widely available – and become cheaper, or even free, to use – they’re being abused by an array of actors, including state actors, to create deepfakes and sow disinformation online.In this session at Disrupt 2024, hear from CCDH CEO Imran Ahmed, CITRIS UC Director of the CITRIS Policy Lab […]
'AI slop' videos may be annoying, but they're racking up views — and ad money
Critics say that "slop" videos made with generative AI are often repetitive or useless. But they get millions of views — and platforms are grappling with what to do about them.
Cite as: van Rooij, I. (2025) AI slop and the destruction of knowledge. This week I was looking for info on what cognitive scientists mean when they speak of ‘domain-general’ cognition. I was curio…
Facebook teaches early lesson in unbounded AI risk
Meta’s chatbot rules court third-rail social issues, like luring users into fantasy. Releasing erratic and powerful tech to consumers and beyond is inherently dangerous. Some $120 bln in venture-backed deals this year and trillions of imputed value gloss over a looming backlash.
A unique survey by MIT Media Lab researchers and an international team of colleagues reveals global attitudes about autonomous vehicles and safety issues.
Your Brain on ChatGPT: Accumulation of Cognitive Debt when Using an AI Assistant for Essay Writing Task – MIT Media Lab
This study explores the neural and behavioral consequences of LLM-assisted essay writing. Participants were divided into three groups: LLM, Search Engine…
Today is a public holiday in Britain, so we bring you a special panel Francis Dearnley hosted during his trip to Germany at the LANDEURO conference hosted by the Association of the U.S. Army. Titled “Ukrainian Innovation at the Speed of Relevance”,
🚀 We’re entering the Agentic Web era
We're witnessing the death of the “search → click → repeat” model.
➡️It’s not just smarter chat. It’s a full-blown machine-to-machine internet, where:
- Agents discover each other based on semantic capabilities, not hardcoded APIs - this is a bit heavy, lighten it up , easier to understand
- Hyperlinks are replaced by inter-agent communication protocols (like MCP & A2A)
- Webpages evolve into active agents with planning, memory, and tool access
- Value creation shifts from user attention to agent invocation and service orchestration
➡️Here’s what happens behind the scenes
▪️User Interaction → You give a high-level intent: “Plan a 3-day trip to Tokyo.”
▪️Plan → An agent interprets your goal, breaks it down, and kicks off orchestration.
▪️Agent Discovery → It finds other agents via registries based on capabilities, trust, or task relevance.
▪️Recruit Agents → Each sub-task (weather, bookings, maps, etc.) gets a specialist agent.
▪️Inter-agent Discussion → They collaborate, negotiate, share outputs, with zero user micromanagement.
▪️Action → The agents execute the task: API calls, reasoning, retrieval, tool use.
▪️Report Result → A polished outcome is returned. You review, tweak, or approve.
▪️Loop → Agents learn, refine, and continue if needed.
➡️What changes in the Agentic Web:
▪️Search becomes delegation → You describe your goal, not keywords.
▪️UX becomes orchestration → Agents plan and act behind the scenes; no more tabs and toggles.
▪️Webpages become agents → Not just containers of content, but intelligent, interactive entities.
▪️Hyperlinks become protocols → Agents use MCP and A2A to communicate and coordinate.
▪️Monetization shifts → It’s no longer about clicks or impressions, it’s about which service an agent chooses to call.
➡️Under the hood:
- Protocols like MCP and A2A enable persistent, asynchronous, and semantically-rich interaction.
- Agents reason using in-parameter LLM knowledge + real-time tool access.
- They coordinate via dynamic capability discovery, workflow negotiation, and task deconstruction.
Paper: https://hubs.ly/Q03CTHgV0
The big shift? The web won’t be built for us anymore, it’ll be built for them.
What happens when the majority of the web is no longer for humans to read, but for agents to act on?
👉Built an AI tool? Get it featured in our community of 12M+
https://hubs.ly/Q03CTHCp0
#agenticai #aiagent #artificialintelligence #webagents #machinelearning
How AI is being used in police stations to help draft reports | CNN Business
Police departments across the country are adopting an AI tool called Draft One, which helps officers write the first draft of police reports. But there are concerns that AI-drafted reports could contain biases or inaccuracies.
We’re about to time travel into the future Sam Altman is building… Subscribe for more optimistic science and tech stories.I interviewed Sam Altman, the CEO o...
The new Scrum Guide describes the whole Scrum Team as self-managing. But how do you lead these teams? Do they need leaders? And if so, where should we leaders focus on? Peter Koning - author of Agile Leadership Toolkit - gives three practical tips
China pitches global AI governance group as the US goes it alone | CNN Business
China has proposed a global action plan to govern artificial intelligence, just days after the United States unveiled its own plan to promote US dominance of the rapidly growing field that’s become a key bargaining chip in trade talks between the economic powerhouses.
