In partnership with

Hello again, human brain, and welcome back to your daily munch of AI news.

Here’s what’s on the menu today:

• AI consciousness is a lie🧠

  Roger Penrose debunks an AI myth.

• WebMD dies while Reddit thrives📈

  How AI affects SEO traffic.

  New here? Subscribe! 😎

Peep today's Spot the AI at the bottom. 👇

MAIN COURSE

Nobel winner crushes AI dreams🧠

Nobel Prize-winning physicist Sir Roger Penrose just debunked the most popular AI myth. He explains why AI systems will never achieve true consciousness, using Gödel's theorem as proof.

What’s the argument?

Penrose argues that consciousness requires understanding, not just computation. While AI can process data and follow rules incredibly well, it doesn't understand why those rules work.

What is Gödel’s theorem?

It proves humans can understand mathematical truths that computers fundamentally can’t. Our ability to “see” why something is true, rather than just following procedural steps, demonstrates that human consciousness operates beyond computation.

Why?

“It doesn't know what it's doing,” Penrose states. AI appears intelligent because modern computers are extremely powerful, but they're still limited to computational mathematics. This is a tiny subset of what conscious minds can grasp.

Will AI ever be conscious?

Penrose believes consciousness must involve non-computable physics that we haven't yet discovered. He suggests quantum mechanics may play a role, particularly the mysterious “collapse of the wave function.”

What about self-driving cars?

Penrose acknowledges AI might outperform humans at specific tasks like driving, but that doesn't indicate consciousness. He compares it to a thermostat that turns off lights automatically without being aware of what it’s doing.

Drowning In Support Tickets? Maven AGI is here to help.

Maven AGI platform simplifies customer service by unifying systems, improving with every interaction, and automating up to 93% of responses. Seamlessly integrated with 50+ tools like Salesforce, Freshdesk, and Zendesk, Maven can deploy AI agents across multiple channels—text, email, web, voice, and apps—within days. Companies like Tripadvisor, ClickUp, and Rho slash response times by 60%, ensuring quicker support and exceptional customer satisfaction. Don’t let support tickets slow you down

Request a free personalized demo today.

SIDE SALAD

AI massacres your site📉

New data reveals how AI search is crushing some content sites. Others are leaning into AI search and thriving in the face of disruption.

Who's losing?

• WebMD: Traffic crashing as AI delivers instant health answers

• Quora: Visits declining as AI provides faster responses

• Stack Overflow: Developer traffic dropping with AI coding tools

• Chegg: Student help platform getting destroyed by AI assistance

• G2: Software review site seeing massive traffic drop

• CNET: Tech news traffic down 70% from four years ago

Who's winning?

• Reddit: Monthly traffic reaching billions despite AI competition

• Wikipedia: Holding steady with 5+ billion monthly visits

• Substack: Growing as readers seek authentic creator content

Why the difference?

Sites relying on SEO and ad revenue from information-seeking are suffering as AI delivers instant answers without clicks. Meanwhile, platforms offering authentic community experiences or original content are thriving.

What's causing this?

Three key triggers:

1. ChatGPT's launch (Nov 2022)

2. GPT-4's release (Mar 2023)

3. Google's AI Overviews rollout (May 2024)

YOUR DAILY MUNCH

Cool Tool 🛠️

An entirely new way to present ideas

Gamma’s AI creates beautiful presentations, websites, and more. No design or coding skills required. Try it free today.

Think Piece 🧠

The key to cheaper, faster, AI services? Chain of Draft (CoD) solves tasks using very concise reasoning steps, reducing words by up to 92% with no loss in accuracy.

Startup News 💰

Anthropic partners with U.S. DOE National Labs. The goal is to put on a “1,000 Scientist AI Jam” to test its new Claude 3.7 Sonnet model on real-world research tasks.

Research 👨‍🔬 

R2-T2 — re-routes mixture-of-experts during test time to fix suboptimal routing. It outperforms bigger models by imitating successful tasks, all without extra training.

NeoBERT — combines modern architecture, updated data, and extended context to outperform older encoders like BERT, RoBERTa, and ModernBERT on MTEB.

MEME FOR DESSERT

SPOT THE AI

3 of these are real penguins. 1 is fake. 🐧

Which one is AI-generated? 👇

Ideas? Comments? Complaints?

Respond to this email or hit me up on 𝕏.

Until next time 🤖😋🧠