Comments URL: https://news.ycombinator.com/item?id=47821856

Points: 3

# Comments: 0

Comments URL: https://news.ycombinator.com/item?id=47304860

Points: 1

# Comments: 0

There’s been a lot of debate about whether Google’s AI Overviews and tools like ChatGPT are actually harming publishers. One publicly traded company’s timeline is worth looking at: Chegg.

What happened (with sources):

2021: Chegg launched Uversity, a platform for educators to share academic content. (Wikipedia)

2023: ChatGPT emerged as a serious competitor in homework help. Chegg responded by launching CheggMate, its own AI product built on OpenAI’s models. (Wikipedia)

Late 2024: Chegg reported accelerating subscriber declines, widely attributed to users shifting to free AI tools instead of paid study platforms. (WSJ, company filings)

Feb 2025: Chegg sued Google, alleging that AI Overviews reduced traffic to Chegg by answering questions directly in search results, harming acquisition and revenue. (Search Engine Land, Reuters)

May 2025: Chegg laid off ~22% of its workforce (≈248 employees), citing competitive pressure from AI and changes in search behavior. (Reuters)

Oct 2025: Chegg announced another round of layoffs (~45%, ≈388 employees), explicitly referencing “the new realities of AI” and reduced traffic from Google to content publishers. (Reuters / SF Chronicle)

What the data suggests (more broadly):

Independent studies show that when Google AI Overviews appear, users are significantly less likely to click through to external sites.

“Zero-click” searches (where users get answers directly on the results page) have increased, especially for informational and educational queries.

The impact isn’t uniform — some publishers report minimal effects — but content that answers how-to, homework, or factual queries appears most exposed.

Why this matters:

Chegg isn’t a small blog or SEO-driven site. It’s a public company with audited financials, legal disclosures, and incentives not to exaggerate under scrutiny. Its filings and lawsuit don’t claim AI is “bad” — they claim that traffic flows are structurally changing.

This doesn’t prove AI search is “killing the web,” but it does show:

AI answers are substituting clicks, not just competing for them.

Entire business models built on informational content are under pressure.

“Build better content” may not be sufficient when answers are synthesized upstream.

Curious how others here see it:

Is this a temporary transition problem?

Or are we watching the unbundling of the open web’s traffic economy in real time?

I’ve been working on ADK-Rust, an open-source framework for building and deploying AI agents in Rust.

The motivation came from building agent systems where performance, safety, and predictable behavior mattered more than rapid prototyping. Most agent frameworks and workflow tools today are Python- or JS-first and tend to be runtime-heavy when taken to production.

Recently, I added ADK-Studio — a visual, low-code environment for building AI agent workflows on top of ADK-Rust.

You can think of ADK-Studio as a Rust-native alternative to tools like n8n, but focused specifically on AI agents: - Visual, drag-and-drop workflow design (sequential, parallel, loop, router agents) - Tool integration (functions, MCP servers, browser automation, search) - Real-time execution with SSE streaming and event traces - Code generation: visual workflows compile down to production Rust code - Build and run agents as native executables directly from the studio

The goal is to let people prototype agent workflows visually, then ship them as fast, memory-safe Rust binaries instead of long-running JS/Python services.

Making AI Agents with ADK Studio is super simple:

1. ADK-Studio install: `cargo install adk-studio` 2. Start ADK Studio server: `adk-studio --port 6000` 3. Open in browser: open http://localhost:6000

I would really appreciate feedback from folks building agent systems, workflow engines, or AI inference infrastructure — especially around design tradeoffs vs existing tools like n8n.

Project site: https://adk-rust.com GitHub: https://github.com/zavora-ai/adk-rust

Best James

Comments URL: https://news.ycombinator.com/item?id=46430090

Points: 3

# Comments: 0

I'm excited to share ADK-Rust - a production-ready implementation of Google's Agent Development Kit in Rust.

Why Rust? After working extensively with adk-python in developing an ai agent factory at zavora.ai, I wanted to bring the same powerful agent development patterns to the Rust ecosystem, targeting use cases where:

Performance is critical - Rust's zero-cost abstractions and memory safety Deployment size matters - Single binary with no runtime dependencies Systems-level integration - Embedded systems, edge computing, IoT Concurrency at scale - Rust's async/await with tokio Features ADK-Rust maintains API parity with the Python ADK where possible:

Model-agnostic - Gemini, OpenAI, Anthropic, DeepSeek support Multiple agent types - LlmAgent, SequentialAgent, ParallelAgent, LoopAgent Tool support - Built-in tools (Google Search, Code Execution) + custom tools MCP support - Model Context Protocol integration Sessions & Memory - InMemorySessionService, DatabaseSessionService Streaming - Full streaming support for real-time responses Telemetry - OpenTelemetry integration for tracing/metrics A2A Protocol - Agent-to-Agent communication

Quick Example

use adk_rust::prelude::*;

#[tokio::main] async fn main() -> Result<()> { let agent = LlmAgentBuilder::new() .name("my_agent") .model(GeminiModel::new("gemini-2.0-flash")?) .instruction("You are a helpful assistant.") .build()?;

    let response = agent.run("Hello!").await?;
    println!("{}", response);
    Ok(())

Links Crates.io: https://crates.io/crates/adk-rust Docs: https://docs.rs/adk-rust Website: https://adk-rust.com/ GitHub: https://github.com/zavora-ai/adk-rust Looking for Feedback I'd love to hear from the community:

What agentic features would you prioritize? Any interest in contributing or testing? Use cases where a Rust implementation would be valuable? This is an independent community project, not officially affiliated with Google, but designed to be compatible with the ADK ecosystem.

Thanks for reading!

Comments URL: https://news.ycombinator.com/item?id=46252336

Points: 2

# Comments: 0

Facing financial challenges and nearing bankruptcy, I decided to document my ideas in a book and sell it on Amazon KDP. The first book, "The Complete LangGraph Blueprint - Build 50+ AI Agents for Business Success", sold over 300 copies in just four weeks and became a best new release in December after being primarily marketed on HN. That reception gave me hope, and I’m deeply grateful to the HN community for your support!

Encouraged by this, I published another book, "People VS AI: The Radical Business Case for Artificial Intelligence Adoption", which also hit #1 in the "Industrial Business Management" category on Amazon (it's still hanging up there on the top spot!)

Now, I’m excited to share my third book: "The Complete Hugging Face Blueprint: A Hands-On Guide to Developing and Deploying Models on Hugging Face with 150+ Practical Lessons." It’s FREE for the next two days (save $49) to encourage your feedback and reviews: Grab it here: https://www.amazon.com/Complete-Hugging-Face-Blueprint-Hands...

This book focuses on practical machine learning without overwhelming theory or math. A key highlight is the Hugging Face Inference API, which I’ve dedicated an entire chapter to. From the feedback I have received so far, it seems few developers know about the Hugging Face Inference API. And they find it amazing when they discover it. I have dedicated a whole chapter in the book to this topic and gladly sharing this below:

What is the Hugging Face Inference API? The Hugging Face Inference API is a user-friendly way to access pre-trained models on the Hugging Face Hub for tasks like: Text classification, summarization, translation, Q&A, Image classification, object detection, Audio transcription and much more.

Key Features: No Deployment Needed: Direct access to thousands of models. Scalability: Handles high-volume requests in real-time. Multi-Framework Support: Works with TensorFlow, PyTorch, and JAX models. Enterprise-Grade Security: Reliable and secure. Low Resources: Does not require GPUs or expensive hardware.

Common Use Cases: Integrating AI into web/mobile apps. Prototyping ML solutions. Real-time predictions for text, image, and audio tasks.

Setting Up the Inference API Step 1: Get an API Token Log in to Hugging Face. Navigate to your "Access tokens" page under settings and generate a token. Step 2: Install the python client pip install huggingface_hub Step 3: Authenticate from huggingface_hub import login login("your_api_token") Example: Sentiment Analysis from huggingface_hub import InferenceClient

client = InferenceClient(model="distilbert-base-uncased-finetuned-sst-2-english") response = client.text_classification("I love using Hugging Face APIs!") print(response) Example: Image Classification client = InferenceClient(model="microsoft/resnet-50") response = client.image_classification("./example.jpg") print(response)

I’ve included over 20 practical examples for real-life use cases in the book. If you’re curious about leveraging Hugging Face or learning Machine Learning through building practical ML applications, I’d love for you to check it out, learn, and share your review! Your support not only helps me improve but also fuels my dream of building Zavora.ai, an AI Agent startup. Thanks so much, HN, for being an amazing community!

Grab the book: https://www.amazon.com/Complete-Hugging-Face-Blueprint-Hands...

Best, James Karanja

Comments URL: https://news.ycombinator.com/item?id=42619160

Points: 1

# Comments: 0

If your boss or co-founder is into tech, AI Agents, or just loves a thought-provoking read, you’ve got to check out People vs AI on Amazon. It’s the ultimate Christmas gift – a deep dive into how AI is reshaping businesses, job roles, and decision-making.

This book is perfect for anyone budgeting for AI Agents in 2025 or just curious about how AI will impact their industry. Whether they’re an exec, a founder, or just someone fascinated by AI’s potential, this book delivers insights they’ll appreciate.

Here’s the kicker: the Kindle version is FREE for the next two days (yes, free!). If you’re like me and love the feel of a real book, the paperback is just $19.99. A thoughtful and affordable gift for any tech enthusiast—or even yourself (because why not?).

Grab it here: https://www.amazon.com/dp/B0DPN1YHFW

Merry Christmas, HN!

Comments URL: https://news.ycombinator.com/item?id=42432180

Points: 3

# Comments: 1

Grab the book: https://www.amazon.com/dp/B0DP69QV7K

Comments URL: https://news.ycombinator.com/item?id=42335821

Points: 3

# Comments: 2