Newsletter Subject

AI Overviews: the publisher predator we never saw coming

From

techinasia.com

Email Address

newsletter@techinasia.com

Sent On

Mon, Jul 8, 2024 02:03 AM

Email Preheader Text

In The Prompt’s third issue, we look at AI overviews targeting media firms and the need for met

In The Prompt’s third issue, we look at AI overviews targeting media firms and the need for metacognition in generative AI. [Read from your browser]( The Prompt 🤖 Welcome to the second edition of The Prompt! Delivered every fortnight, this free newsletter breaks down the biggest stories and trends in AI. Get all our content, including The Prompt, by [registering for a Tech in Asia account](. Alternatively, [sign up here]( if you just want to receive this newsletter. IN FOCUS In today's newsletter, we look at: - [Error 404: Web not found, killed by Google’s AI search]( - [Singapore's data center crunch drives AI focus]( - How merely scaling up AI models won’t lead to artificial general intelligence --------------------------------------------------------------- Hello {NAME} I distinctly recall attending a weekly meeting at my previous company in May 2023. We used these meetings to discuss new ideas, recent readings, and industry developments. The topic of the day was the [AI Overviews]( which might have an impact on our business as a media company. We discussed how AI Overviews by search giants like Google or Bing would essentially be able to summarize our articles, offering readers the option to get the gist of an article without clicking on it. This would impact the revenue model of publishers who relied on advertisements. I like to think we were among the first to predict that companies like Google might also turn AI Overviews into another ad opportunity, which could further diminish the organic reach of news websites. A year later, this prediction has become a reality. In this week's Big Story, my colleague Scott writes about the pain points media industry professionals are experiencing. Crucially, the presence of AI Overviews could result in a [25% reduction]( in web traffic over the next year. Publishers, search engine optimization experts, and marketers are unhappy about having to rely on one or two channels (usually Google and Meta) for their web traffic, and now AI Overviews are set to take away more of these publishers' traffic. How then should media outlets navigate this change? AI search, according to editorial expert Andrew Thompson, “will reward longer-form content, more unique content, and content that comes from original expertise and experience.” Moving on to this week’s Hot Take, I take a closer look at recent discussions on metacognition in generative AI. This comes in light of Bill Gates speaking up about the need for implementing metacognition, instead of merely scaling up current AI models. -- Lokesh  --------------------------------------------------------------- THE BIG STORY [Error 404: Web not found, killed by Google’s AI search]( Google’s AI Overviews marks a major overhaul of how people search the internet, but there are concerns about how it will impact various businesses. --------------------------------------------------------------- THE HOT TAKE Why achieving AGI requires more than just bigger models Here’s what happened: - On June 29, Bill Gates discussed the concept of "metacognition" on the [Next Big Idea]( podcast. - Metacognition involves a system's ability to think about its own thinking process. - Gates highlighted that current AI models like GPT-4 lack sophisticated cognitive strategies. Here’s our take: At the World Economic Forum [earlier this year]( OpenAI CEO Sam Altman said that artificial general intelligence (AGI) could be developed in the “reasonably close-ish future.” AGI refers to AI systems that [exceed the intelligence of humans](. This may seem far-fetched, especially since ChatGPT is still so prone to hallucinations. Perhaps, since AGI’s development requires large sums of investments, tech startups have figured out that the best way to boost their valuations is to tease that AGI is imminent. AI startups globally raised [more than US$27.1 billion]( between April and June this year, showing that investors are buying into the story. Currently, generative AI models produce output on the basis of [patterns and trends]( identified from analyzing vast amounts of data, which requires large amounts of brute computing power. However, human thinking isn’t just about interpreting data. When faced with a problem, for instance, a person should take a step back and evaluate how best to go about solving the problem - instead of responding right away. At Tech in Asia, we’ve had some experience of how complicated it can be to work with genAI. When we were experimenting with news automation, to get a satisfactory response from our chatbot, we had to prompt it with at least 30 action items, mimicking how a journalist would think about writing a piece. The idea of metacognition brings fresh air to the talk about building ever-larger models with ever-increasing computing power. Achieving metacognition in genAI involves creating feedback loops that enable continuous monitoring and evaluation of outputs. Techniques such as [self-reflection]( in which the AI analyzes past decisions and outcomes, can significantly enhance future performance. Additionally, estimating uncertainty and generating confidence scores allow the AI to recognize its limitations. Imagine an AI model that can understand context, reflect on its thoughts, and make decisions based on deeper reasoning. This could bring the benefits promised by the promoters of genAI, but without the drawbacks - hallucinations, out-of-this-world power consumption, etc. Of course, this is easier said than done. It requires a shift in focus from just scaling models to enhancing their cognitive abilities. In the end, the race to AGI isn't just about who can build the biggest model. It's about who can build the smartest one. But if investors seem happy to just scale the current technology, would companies be interested in innovating?  --------------------------------------------------------------- DEEP READS Check out these interesting reads about AI: 1️⃣ [OpenAI’s ChatGPT Mac app was storing conversations in plain text]( Until June 28, OpenAI's ChatGPT macOS app had a security flaw: Local chat records were stored in plain text, so anyone with access to the Mac device could read them. However, OpenAI quickly released an update to encrypt the chats. 2️⃣ [Prying open the AI black box]( AI is changing biology by helping analyze large amounts of genome data to find new treatments. However, understanding how the tech makes these discoveries can be hard. To help, scientists at the Cold Spring Harbor Laboratory in the US created SQUID, a tool that explains how AI models work in genomics. 3️⃣ [Can the climate survive the insatiable energy demands of the AI arms race?]( The AI boom has boosted big tech stocks but is hindering climate goals because AI data centers have increased greenhouse gas emissions. Companies like Google and Microsoft are striving for sustainability but fear missing their 2030 net-zero targets.  --------------------------------------------------------------- FYI 1️⃣ [Singapore's data center crunch drives AI focus]( Tech firms will likely leverage their presence in Singapore for AI workloads, while shifting more traditional workloads to Malaysia and Indonesia. That’s it for this edition - we hope you liked it! Do also check out previous issues of the newsletter [here](. See you next time!  --------------------------------------------------------------- EVENTS HAPPENING You can also check out a curated list of trending tech events [over here]( and Tech in Asia’s signature events [here](. [Tech in Asia Conference Kuala Lumpur 2024 : Malaysia on The Rise on July 24 - 25Â]( 1,000 founders and investors have already signed up, and you're invited. This is your chance to be part of the dynamic Southeast Asian tech scene. Hear from top industry speakers like Nadira Yusoff, CEO of Kiddocare, Sheyantha Abeykoon, Group CEO of Boost and Kel Jin Chua, Director of Shopee Singapore at the stage sessions of #TIAConferenceKL. Final release tickets are selling fast. [Secure your tickets today.Â]( [Tech in Asia’s Founders Meetup in Malaysia on July 25]( Founders, clear your calendar for July 25. We're bringing together Southeast Asia's most influential investors and startup leaders for an evening of game-changing conversations, exclusive connections, and actionable insights designed to help you scale your startup and dominate the region’s tech scene. [Get your tickets now]( - your future self will thank you. [KL Conference Exclusive: Networking Night on July 24]( Seize the opportunity to connect with the brightest minds and most influential players in Asia's tech ecosystem, all while enjoying panoramic city views and delicious treats. This exclusive event is only open to Tech in Asia Conference Kuala Lumpur Pro Pass holders. [Secure your spot on our guest list today!]( [KL Conference Exclusive: After Party on July 25]( After two days of high-impact learning, join us at the official closing party to unwind, socialize, and celebrate the thriving startup scene in Malaysia. There’s no better way to do so than by immersing yourself in its vibrant energy firsthand. This exclusive event is only open to Tech in Asia Conference Kuala Lumpur Pro Pass holders. [RSVP now to secure your spot!Â]( [Tech in Asia Conference Jakarta on Oct 23-24]( Where is the next Silicon Valley? It might just be Jakarta! Indonesia's young tech talent is booming, and Tech in Asia Conference Jakarta puts you right in the center of the action. Join us at #TIAConferenceJKT to connect with 4000+ region's brightest minds, 150+ hottest startups, and hear straight from 80+ experts. Early-bird tickets are now available! [Snatch the 60% discount while it lasts!]( --------------------------------------------------------------- That’s it for this edition - we hope you liked it! Not your cup of tea? You can unsubscribe from this newsletter [here](. In the meantime, if you have any feedback or ideas, feel free to get in touch with Terence, our editor-in-chief, at terence@techinasia.com. See you soon! P.S. Don't miss out on the biggest tech news and analysis. Add newsletter@techinasia.com to your address book, contacts, or safe sender list. Or simply move us into your inbox. Too many emails? Switch to a different frequency or get new content through our [preference center](. You can also break our hearts and remove yourself from all Tech in Asia emails over [here](  Copyright © 2024 Tech in Asia, All rights reserved. 63 Robinson Road, Singapore 068894

Marketing emails from techinasia.com

View More
Sent On

08/12/2024

Sent On

06/12/2024

Sent On

09/11/2024

Sent On

03/11/2024

Sent On

30/10/2024

Sent On

28/10/2024

Email Content Statistics

Subscribe Now

Subject Line Length

Data shows that subject lines with 6 to 10 words generated 21 percent higher open rate.

Subscribe Now

Average in this category

Subscribe Now

Number of Words

The more words in the content, the more time the user will need to spend reading. Get straight to the point with catchy short phrases and interesting photos and graphics.

Subscribe Now

Average in this category

Subscribe Now

Number of Images

More images or large images might cause the email to load slower. Aim for a balance of words and images.

Subscribe Now

Average in this category

Subscribe Now

Time to Read

Longer reading time requires more attention and patience from users. Aim for short phrases and catchy keywords.

Subscribe Now

Average in this category

Subscribe Now

Predicted open rate

Subscribe Now

Spam Score

Spam score is determined by a large number of checks performed on the content of the email. For the best delivery results, it is advised to lower your spam score as much as possible.

Subscribe Now

Flesch reading score

Flesch reading score measures how complex a text is. The lower the score, the more difficult the text is to read. The Flesch readability score uses the average length of your sentences (measured by the number of words) and the average number of syllables per word in an equation to calculate the reading ease. Text with a very high Flesch reading ease score (about 100) is straightforward and easy to read, with short sentences and no words of more than two syllables. Usually, a reading ease score of 60-70 is considered acceptable/normal for web copy.

Subscribe Now

Technologies

What powers this email? Every email we receive is parsed to determine the sending ESP and any additional email technologies used.

Subscribe Now

Email Size (not include images)

Font Used

No. Font Name
Subscribe Now

Copyright © 2019–2025 SimilarMail.