In partnership with

Yesterday, a 286-page ruling was released that could shake up the multi-billion-dollar search industry.

Google, with its search business generating a staggering $175 billion annually, controls 90% of global search traffic—more than the GDP of some countries.

Google’s wealth primarily comes from its ad network. It controls:

• 80% of ad market

• 91% of publisher ad services

• 50% of ad exchanges

Surprisingly, Mozilla is funded by Google to maintain its default search engine status. Mozilla receives funding from several places, but Google remains the largest contributor.

The reason? It’s part of Google’s strategy to dodge monopoly accusations. The real surprise? Google pays $18-20 billion a year to Apple just to remain the default search engine on its devices.

Google’s dominance isn’t just confined to its platform. A whopping 95% of premium searches come from Apple devices, and this is why Google forks out billions to ensure it stays as the default search engine on iPhones and iPads. Samsung, too, benefits from this hefty payment.

Google has been prioritizing short-term ad revenue at the expense of search quality. As a result, the search experience has deteriorated, often leading users to irrelevant or less useful results.

For instance, I sometimes find myself using Google just to locate Reddit threads, rather than directly getting quality answers. This decline is particularly frustrating for users who rely on the platform for efficient and relevant search results.

The EU Antitrust regulators also running after Google for the similar charges.

Then whats next?

In 1982, the Bell System was broken up into smaller parts like AT&T, Verizon, and others. Now, there’s increasing speculation that Google (or its parent company, Alphabet) could be forced to split into divisions like Chrome, YouTube, Android, and others to avoid monopoly scrutiny.

Will this breakup happen? The next few months may offer a clearer view of how the tech giant navigates this monumental challenge.

By the way there is a comic relief, there is character called “Monopoly Man” - Read the reddit :)

🌸 From The Agent Community

Over the past couple of weeks, I have read more than 100 research papers on agents. I noticed that many researchers are focusing on simulating a society of agents and studying how these agents behave.

I have categorized the scenarios into three main groups:

• Society Simulation with LLM-based Agents (E33)

• Behaviour of LLM-based Agents (This Edition - E39) - Part 1

• Environment for Agent Society. (Next Edition)

“We want AI agents that can discover like we can, not which contain what we have discovered.”

🌼 Reflexion: Language Agents with Verbal Reinforcement Learning

A novel method for training agents through linguistic feedback rather than traditional weight updates.

Supports scalar values, free-form language, and both external or simulated feedback sources.

Reflexion achieved 91% pass@1 on the HumanEval coding benchmark, surpassing GPT-4's 80%.

🌼Exploring Collaboration Mechanisms for LLM Agents: A Social Psychology View.

This paper explores how NLP systems can mimic human collaborative intelligence by testing interactions among multiple LLMs.

The authors create four societies, each with distinct agent traits (easy-going or overconfident) and collaboration styles (debate or reflection) - Certain collaboration strategies outperform previous methods and reflect human-like behaviors such as consensus and conformity.

🌼 AgentVerse: Facilitating Multi-Agent Collaboration and Exploring Emergent Behaviors in Agents.

A multi-agent system inspired by human group dynamics that enables LLM-powered agents to collaborate effectively as a cohesive team on tasks such as text understanding, reasoning, coding, and tool utilization.

💥 Use AI to 10X your productivity & efficiency at work (free bonus) 🤯

Still struggling to achieve work-life balance and manage your time efficiently?

Join this 3 hour Intensive Workshop on AI & ChatGPT tools (usually $399) but FREE for first 100 readers.

Save your free spot here (seats are filling fast!) ⏰

Want to be one of these people & be a smart worker?
Free up 3 hours of your time to learn AI strategies & hacks that less than 1% people know! 

🗓️ Tomorrow | ⏱️ 10 AM EST

In this workshop, you will learn how to:

✅ Make smarter decisions based on data in seconds using AI 
✅ Automate daily tasks and increase productivity & creativity
✅ Skyrocket your business growth by leveraging the power of AI
✅ Save 1000s of dollars by using ChatGPT to simplify complex problems 

👉 Hurry! Click here to register (FREE for First 100 people only) 🎁

🌼 ChatEval: Towards Better LLM-based Evaluators through Multi-Agent Debate.

A multi-agent debate-based system that uses LLMs for autonomous evaluation of text, moving beyond single-agent strategies.ChatEval mimics human evaluation processes by enabling multiple LLMs to collaborate, enhancing the reliability and accuracy of assessments.

The system leverages distinct LLM capabilities, making it more efficient in evaluating open-ended questions and traditional NLG tasks.

🌼 Suspicion Agent: Playing Imperfect Information Games with Theory of Mind Aware GPT-4

Leverages GPT-4’s high-order theory of mind (ToM) - designed to excel in imperfect information games, which mirror real-world decision-making under uncertainty.

With proper prompt engineering, Suspicion-Agent shows strong adaptability across various card games, using only game rules and observations.Demonstrates promising results in games like Leduc Hold’em, potentially outperforming traditional algorithms without specialized training.

🌸 Startups

🌼 Selling Sunlight at nights by putting mirrors on space.

In the near future, Reflect Orbital had a bold vision: a way to harness solar energy from space. Their mission was to launch 57 small satellites, each equipped with mirrors capable of reflecting sunlight to Earth. The tech was cutting-edge, though reminiscent of a 1990s Russian project called Znamya, which had a similar dream but a much shorter lifespan—its satellite fell back to Earth after just a few hours.

To prove their concept, Reflect Orbital tested a mirror high above the clouds. Suspended on a hot air balloon 1.7 miles above a solar farm, the mirror caught the light of the sun and directed it toward the panels below. The results were astonishing—500 watts of energy per square meter, about half the brightness of the sun. It will defiantly solve the storage problem of solar electricity.

🌸Choice Cuts

🌼 Large Language Models (LLMs) cannot genuinely explain their past behavior because they lack memory or self-awareness. When asked to explain a previous response, LLMs rely only on the context available to them at that moment and construct a justification based on that.

They don’t retain or recall any thoughts from earlier in the conversation.

Even approaches like “Tree of Thought,” which attempt to track previous states, only record the context. Jonathan, the writer, executed an experiment on generating number patterns.

🌼AI is empowering us to tackle challenges we once thought were out of reach.

While we are worried AI might make people passive, there’s evidence to suggest the opposite: AI might actually make some of us more ambitious.

Yes, chatbot help is fallible, but it’s also making projects that once seemed insurmountable more approachable, enabling us to try things we might not have considered before. In a sense, generative AI offers the potential for users to explore more ambitious avenues in their learning and creativity.

🌸 Podcasts

There’s a lot more I could write about but I figure very few people will read this far anyways. If you did, you’re amazing and I appreciate you!

Love MusingsOnAI? Tell your friends!

📮Want to Advertise with us?

If your company is interested in reaching an audience of AI professionals and decision-makers, reach us.

If you have any comments or feedback, just respond to this email!

Thanks for reading, Let’s explore the world together!

Raahul