Anthropic's bold hire

Google’s AI still biased

AI Tool Report

Welcome to AI Tool Report!

Thursday’s top story: OpenAI rival, Anthropic, has hired Instagram co-founder—Mike Krieger—to lead its product development, as it looks to bring Claude into the workplace.

🎙️ Plus: A new podcast episode is out: Listen as an OpenAI consultant talks about Enterprise upskilling with AI.

🌤️ This Morning on AI Tool Report

  1. 🔥 Anthropic's bold new hire

  2. 🌱 Microsoft's carbon plans unravel

  3. 🤔 Google’s AI still biased?

  4. 🚨 Google spying on calls?

Read Time: 5 minutes

Thursday’s newsletter is in partnership with Osum


Stock tracker


🔥 Anthropic's bold new hire

Anthropic's bold new hir

Our Report: Anthropic announced that the co-founder of Instagram and Artifact (a personalized AI news-reading app, recently acquired by Yahoo), Mike Krieger, will be joining the company as its first Chief Product Officer (CPO).

🔑 Key Points:

  • Krieger will oversee Anthropic’s product engineering, management, and design efforts in a bid to expand their AI apps and bring Claude (its OpenAI equivalent) to a wider audience.

  • Anthropic CEO—Dario Amodei—believes that his background “will be invaluable, as we create new ways for people to interact with Claude, particularly in the workplace.

  • This statement suggests that alongside consumer-focused projects—like the recently released iOS Claude app—Krieger may also be responsible for finding business-focused uses for Claude.

🤔 Why you should care: Anthropic has brought this big-name executive on board just days after its rival OpenAI launched its new AI model (GPT-4o) and Google announced major AI updates at its I/O conference, in a seemingly aggressive push to take on the AI giants.

Together with Osum

Osum ad

Using up-to-date customer insights can result in higher quality leads and contribute to 175% growth in sales with 72% faster lead conversion…

However, market research projects can easily cost 1000’s of $$$ and take weeks to complete.

Osum helps you perform deep market research to identify untapped growth potential in seconds.

Simply copy-paste any URL to instantly get detailed industry insights, competitive analysis, targeted customer personas, sales prospect insights, actionable growth opportunities, and more for any product or business.

Stay ahead of your competition and discover new ways to unlock growth.

P.S. AI Tool Report readers get 25% off on any plan. Use the code AITOOLREPORT at checkout.

Trending Tools banner
  1. Verbalate translates video content into multiple languages

  2. JobHuntMode is an AI-powered job-hunting tool

  3. Documind is for analyzing PDFs

  4. Jumble is an AI journaling tool

  5. Search Alkemy is an SEO keyword research tool


🌱 Microsoft's carbon plans unravel 

Microsoft's carbon plan unravel

Our Report: After agreeing to cut its greenhouse gas emissions, Microsoft’s carbon footprint has grown by 30% since 2020, thanks to its prioritization of AI development.

🔑 Key Points: 

  • In 2020, Microsoft pledged to slash its greenhouse gas emissions by half; capture more CO2 emissions than it produced, and become carbon-negative by 2030.

  • Although they’re ‘100% powered by renewable energy’, this is largely through Renewable Energy Credits (RECs) which don’t produce green energy, they just mask true carbon emissions.

  • Last year, their electricity consumption was higher than a small country’s, and their production of CO2 was equivalent to the whole of Haiti’s annual carbon pollution.

🤔 Why you should care: AI requires huge amounts of data processing, which is done in data centers on computers that consume a lot of energy, and as big tech giants—like Microsoft, OpenAI, Google, and Apple—all compete in the AI race, this presents a wider issue that could affect the global goal of reaching net zero by 2050.

Breaking news banner

🤔 Google’s AI still biased?

Our Report: Despite announcing a plethora of major AI updates at its annual I/O developer conference—including AI search features and a Sora text-to-video equivalent—Google still hasn’t fixed its biased image generator.

🔑 Key Points: 

  • Back in February, Google was forced to pause its AI-powered image generator feature after users complained it was creating inaccurate and offensive images of historical people.

  • At the time, the co-founder of Google DeepMind–Demis Hassabis–said they would fix the issue within a couple of weeks, but we’re now in May, and the image generator is still switched off.

  • Experts believe that the issue was likely caused by some lazy hardcoding, that was initially done to rectify bias when training the chatbot, which they're now having to painfully unpick.

🤔 Why you should care: This raises questions about Google’s approach when building its AI models, and in light of their recent AI announcements, some are concerned about how ethically sound their new AI developments will be.

P.S You might want to read the next story, Google is in more hot water… ⬇️ 


🚨 Google spying on calls?

Our Report: Google demoed an AI-powered scam call detection feature at its I/O developer conference yesterday, aimed at protecting Android users from phone scams, but privacy and security experts are blasting it as ‘dangerous’.

🔑 Key Points:

  • The new feature analyzes conversations, in real-time, and detects fraudulent language patterns associated with scams, suggesting that Google might listen to private phone calls.

  • Meredith Whittaker, ex-Google employee and president of the encrypted messaging app–Signal—warned that this was “dangerous as it's just “a short step from client-side scanning”.

  • Client-side scanning is when a system scans (voice) content against a database of illegal, harmful, or hateful content, and if there’s a match, it fires a message to the recipient.

🤔 Why you should care: Although this feature has been designed to prevent scams, it has the potential to be abused as Google could easily potentially program the technology to scan for other keywords, therefore abusing user privacy.

🍔 Quick hits

  • Humane’s AI pin is now powered by OpenAI’s newest AI model, GPT-4o, and is reportedly now faster and gives better answers.

  • US Senators have proposed a roadmap for regulating AI, calling for $32B for non—defense AI innovation.

  • Google TV is getting Gemini AI-powered content recommendations based on user behavior.

We read your emails, comments, and poll replies daily.

Hit reply and let us know what you want more of!

Until next time, Martin & Liam.

P.S. Don’t forget, you can unsubscribe if you don’t want us to land in your inbox anymore.

What did you think of this edition?

Login or Subscribe to participate in polls.