⚙️ Exclusive interview: SEO expert explains how to navigate AI search

Subscribe | Advertise

Good morning. Nvidia reported — and I quote — “record quarterly revenue of $26.0 billion, up 18% from Q4 and up 262% from a year ago” on Wednesday. The semiconductor also announced a 10-for-1 stock split effective June 7.

To the surprise of no one, Nvidia did it again, and the stock surged in after-hours trading. All this winning must get exhausting for CEO Jensen Huang.

In other news, I sat down with SEO expert Jim Yu to talk about Google’s new world of AI search. Read on for the full story.

In today’s newsletter: 

• 📱Ads are coming to Google’s AI overviews, and organic search will pay the price

• 🛜 Geoffrey Hinton says AI can have feelings; experts disagree

• ◼️ Anthropic cracks open the ‘Black Box’ of AI

• 💻 Exclusive interview: SEO expert explains how to navigate AI search

Ads are coming to Google’s AI overviews, and organic search will pay the price

Last week, Google finally brought generative AI into search with AI Overviews. Now Google — which has maintained a 90% share of the search engine market — is on the verge of bringing ads into this newly AI-fueled ecosystem. 

The details: At its Marketing Live event, Google said that it will soon start testing Search and Shopping ads within AI Overviews (for users in the U.S.). These ads, matched to search queries, will be placed above and below each Overview, further pushing organic links down.

• “Let’s say friends are renovating and they search for ‘short-term storage.’ Clicking an ad for a storage facility may lead to a dynamic experience where AI helps them figure out what they need,” Google said in a blog post.

The big picture: Kristine Schachin, an SEO expert, called the practice “monopolistic.”

• “They created AI Overviews to summarize websites thus STEALING the click. Now they will put ads around the summaries,” she said. “This is monopolistic.” 

The idea of using advertising to monetize generative AI output is one that VC Ari Newman recently called “terrifying.”

Geoffrey Hinton says AI can have feelings; experts disagree

In a recent conversation with Joel Hellermark, the founder and CEO of Sana, Geoffrey Hinton — recognized for his work on neural networks — said that in 1973, he saw a robot express an emotion. 

• The robot could assemble a toy car if the pieces were laid out separately, but its vision wasn’t good enough to assemble the car if the pieces were in a pile; when piled up, the robot would hit the pile to separate the pieces.  

  • To Hinton, feelings are “actions we would do if it weren’t for constraints.”

“If you saw that in a person, you’d say it was cross with the situation,” Hinton said. 

Nope, not really: First, feelings and emotions — according to Dr. Neel Burton — are much more complex than Hinton makes them seem. Burton has said that feelings are a (brief) emotional experience & emotions are the underlying (longer-term) source of that experience. 

• Hinton’s pronouncement was met with derision from a number of machine learning experts on Twitter. 

“Great example of projection – people take feelings and conclusions they make about humans (‘he was cross’) and copy-paste them onto inanimate objects,” AI researcher Dr. Sasha Luccioni said. “That does not emotions make!”

Together with Superhuman

Superhuman is the fastest email experience ever made.

Now with Ask AI, you can immediately get answers from your inbox.

Just ask: “Where is the q2 offsite”, or “summarize recent Deep View newsletters.”

What took minutes now takes seconds.

Stop searching, start asking with Superhuman AI.

Get 1 month free

Anthropic cracks open the ‘Black Box’ of AI

The AI models of today are often referred to as “black box” models. The reason behind this is simple: These types of models are made directly from data by an algorithm. This, according to machine learning expert Dr. Cynthia Rudin, means that “humans, even those who design them, cannot understand how variables are being combined to make predictions.”

• Such systems present obvious ethical concerns — If we don’t know how a model works, or what it was trained on, we can’t say for sure if it is safe.

Anthropic, a leading AI lab, on Tuesday published new research in which its scientists were able to crack the black box of one of its models, gaining a bit more understanding of how the model works. 

• Researchers used a process called “dictionary learning” to extract millions of features — monosemantic components — from the model. 

• This resulted in “a rough conceptual map of its internal states halfway through its computation.”

“If we could really understand these systems — and this would require a lot of progress — we might be able to say when these models actually are safe, or whether they just appear safe,” Chris Olah, the head of Anthropic’s interpretability team who led the research, told Time.

💰AI Jobs Board:

• Generative AI Scientist, Associate: Goldman Sachs · United States · New York · Full-time · (Apply here)

• Business Development Manager, Space Robotics: GitAI · United States · Torrance, CA · Full-time · (Apply here)

• Generative AI - Generalist: Surge AI · United States · San Francisco, CA · Full-time · (Apply here)

 🔭 Tools: *

• Enquire PRO: This human + AI platform helps you save time and money, while solving big problems at work.

  • By empowering you to connect and collaborate with the right experts in any industy or location, you can make better decisions, faster.

🌎 The Broad View:

• Here’s how gas mileage calculations really work (Vox).

• A Wisconsin software engineer was arrested for creating/distributing CSAM with AI (The Verge).

• TikTok is planning to cut a big chunk of its staff (The Information).

*Indicates a sponsored link

Together with Zaplify

Need more prospects? Try this new AI Sales Assistant

Spam filters. Irrelevant LinkedIn InMail templates. Obvious automated follow-ups that don’t work.

If you’re ready to ditch the {insert_first_name} tags for real results, try Zaplify’s new AI Sales Assistant.

Pairing human touch with a savvy AI, this tool helps find your next prospect, engage meaningfully, and retain relationships – all in one platform.

Start by searching from 450 million people, and the AI Sales Assistant will…

• Suggest similar people worth pursuing

• Draft hyper-personalized messages based on prospect’s profile/background

• Revive old leads and make you reach a new level of reply rates

Plus, it unifies your LinkedIn and email chats so you never lose a prospect again.

Join the waitlist and be the first to experience the future of authentic outreach.

Exclusive interview: SEO expert explains how to navigate AI search

Even as Google is figuring out ways to make as much money as it possibly can from its coming evolution of Search, publishers and companies are trying to figure out the whole new world of the internet that Google just ushered in (did somebody say market power?). 

I sat down with Jim Yu, the founder of veteran enterprise SEO company BrightEdge, and here’s the first thing he told me: “I don’t think the internet as we know it is dead. But it’s changing.” 

A little different, a little the same: Google’s been experimenting with “zero-click results” for years now in the shape of those little drop-down menus that crop up throughout a search page. Yu said AI summaries are like that, “but on steroids.” 

• The big difference with the AI Overviews, however, is that they offer an “informed opinion” on a given search query. That, Yu said, is a “fundamental” shift in the world of search engines, which historically have been neutral purveyors of information. 

The world of AI Search: Some search engines, like Perplexity, offer citations in their AI-generated output. Yu said that these citations will become the new form of ‘ranking.’ He added that BrightEdge has found that 50% of these citations were also listed on the front page of normal Google. 

• AI search engines “still have to have sources of original information,” Yu said. “It still has to derive authority, it still has to derive trust.”

The bright side: The story of AI in search — for businesses, at least — isn’t all bad, according to Yu. Overviews will send the value per potential shopper way higher than it is currently: “By the time you reach that end site, you are darn near ready to transact.” 

• The offshoot of this reality is that it could very likely beat down smaller merchants as Google seeks to position itself as a threat to Amazon. 

Privacy as a business: At the same time, I’ve spoken to a number of people — and seen a lot of posts like the one below — who would prefer a search engine whose hallmark is privacy, security and no generative AI. I asked Yu if he thinks there’s a chance that Google’s jump into AI search will break its dominance, opening the door for smaller search engines with those attributes to fill the gap in consumer demand. 

“There’s definitely clear parts of the market where privacy will be number one, namely for businesses and enterprises,” he said. “For end consumers, is privacy sharp enough on a single product basis to sustain as a business model? That I don't know.” 

And on that, I suppose we’ll see.

• Seodity: An AI-powered SEO tool.

• iListen: A tool to summarize articles & text into miniature podcasts.

• Reflect: An AI-powered note-taking tool designed to mirror the way we think.

Have cool resources or tools to share? Submit a tool or reach us by replying to this email (or DM us on Twitter).

*Indicates a sponsored link

SPONSOR THIS NEWSLETTER

The Deep View is currently one of the world’s fastest-growing newsletters, adding thousands of AI enthusiasts a week to our incredible family of over 200,000! Our readers work at top companies like Apple, Meta, OpenAI, Google, Microsoft and many more.

If you want to share your company or product with fellow AI enthusiasts before we’re fully booked, reserve an ad slot here.

One last thing👇

That's a wrap for now! We hope you enjoyed today’s newsletter :)

We appreciate your continued support! We'll catch you in the next edition 👋

-Ian Krietzberg, Editor-in-Chief, The Deep View