AI Face-off: France vs. Silicon Valley

News, AI leaders, business insights and more

This Week’s Line-up

  • AI Face-off: France vs. Silicon Valley

  • Perplexity: Love, Lies, & AI

  • Trend Alert: AI and the Future of Gaming

  • Woman in AI: Meet Rana el Kaliouby

  • Case Study: Toys "R" Us Debuts AI-Made Ad

  • Pro Tip: How to Define a Winning AI Use Case

ps. Connect with me on LinkedIn and Instagram for daily news, insights, and more. 😎

NEWS YOU CAN’T MISS

AI Face-off: France vs. Silicon Valley

French AI lab Kyutai, backed by billionaire Xavier Niel, introduces Moshi, a cutting-edge AI voice assistant generating significant buzz in the tech world and challenging OpenAI's ChatGPT.

Moshi, the "first real-time voice AI assistant," boasts 70 emotions, human-like interactions, and thinks as it speaks with an impressive 160ms latency.

I tried Moshi. It's really good and fast … but it’s not as engaging or comprehensive in conversations as OpenAI's ChatGPT-4o's voice chat. But that could change quickly as Moshi evolves from prototype to product.

Here's why I find Moshi so interesting:

  • Efficiency: It took only 8 people and 6 months to build Moshi. 🤯

  • Funding: It required $324 million in funding, a stark contrast to the billions invested in OpenAI.

  • Open Source: Non-profit Kyutai will make Moshi’s code open source, enabling anyone from Bangalore to Rio to Lagos to create their own voice assistants.

French AI Leadership: Moshi joins Mistral and French-origin startup Hugging Face, further solidifying France's position as a major player in the global AI landscape. 🇫🇷

ps. For those who claim innovation dies in Europe due to regulations like the EU AI Act, France seems to be proving otherwise: responsible AI governance and innovation can indeed coexist.

Perplexity: Love, Lies, & AI

Perplexity co-founders Johnny Ho, Aravind Srinivas and Denis Yarats. Source: Perplexity

I loved using Perplexity. Correction: used to love using Perplexity.

Quick recap: Perplexity is an AI-powered search engine that answers questions conversationally, like ChatGPT, but using real-time web data. This makes it ideal for research that requires the most current information.

Despite being hailed as a potential Google competitor, Perplexity is now under fire, facing legal action from publishers over accusations of plagiarism and copyright infringement.

Here's a breakdown of the allegations:

  • Plagiarism: Forbes accused Perplexity of stealing content and creating "knockoff stories" using "eerily similar wording" and "entirely lifted fragments" from its articles.

  • Unauthorized Web Scraping: Wired reported that Perplexity was surreptitiously scraping content from websites that had explicitly blocked its crawler.

  • Fabrication: Wired dubbed Perplexity a "bullshit machine" due to instances where it generated completely fabricated information.

This hasn't deterred investors, though. Softbank's Masayoshi Son is reportedly considering an investment that would raise Perplexity's valuation to $3 billion.

As for me, I still use Perplexity sometimes because it's the only generative AI search engine with real-time info besides Gemini. But I'm more cautious now, and a bit guilty. What if it's using unattributed work?

Is this the inevitable cycle of AI innovation: love, disillusionment, and then a reluctant compromise?

TREND ALERT

AI and the Future of Gaming

Image: Inworld AI

Are you a gamer? If you’re not, you might be shocked at the statistics:

  • Over 40% of the world – a mind-blowing 3.32 billion people – are active video gamers. 🤯

  • Asia leads the pack with 1.5 billion gamers, and men dominate the scene, making up 53% of players.

Gaming is big business and AI is coming to it big time.

According to a report by Unity, 62% of game development studios now utilize AI tools at some stage of development. Two ways in which AI is being used:

  • Content Generation: Imagine vast, intricate game worlds, unique characters, and captivating quests, all created automatically by AI. This means more variety, endless replayability, and shorter development times.

  • Intelligent NPCs: Say goodbye to predictable, robotic characters. AI-powered non-player characters (NPCs) are becoming lifelike, adapting to your actions and surprising you with realistic behavior.

Here’s a quick look at two companies:

Inworld AI:

  • Develops AI-powered virtual characters with advanced conversational abilities and emotional depth. Their answers and actions are unexpected and creative, which makes the game more uncertain and engaging.

  • Announced a partnership with Xbox and Microsoft and has raised $50 million at a $500 million valuation.

Roblox:

  • Building upon its existing tools, Roblox is now setting its sights on 4D experiences – the dynamic interplay between people, objects, and environments. FYI, 77 million people play Roblox daily.

Ok I admit, this sounds like fun. Are you ready?

WOMEN IN AI

Meet Rana el Kaliouby

Celebrating this week's woman in AI 🥳: Meet Dr. Rana el Kaliouby, a pioneer in the field of Emotion AI, which focuses on developing technology that can understand and interpret human emotions.

Born in Egypt, el Kaliouby earned her bachelor's and master's degrees from the American University in Cairo, followed by a Ph.D. from the University of Cambridge and a post-doctorate at MIT Media Lab.

In 2009, el Kaliouby co-founded Affectiva, a groundbreaking startup that decoded human emotions and cognitive states through facial and vocal expressions. The company, used by an impressive 26% of Fortune 500 companies, was acquired by Smart Eye in 2021.

el Kaliouby also published a bestselling book, "Girl Decoded," in 2020, which not only delves into the technical aspects of her work but also addresses the cultural and personal challenges she faced as a Muslim woman in the male-dominated tech industry.

I loved the book and highly recommend it! 🥇

She has received numerous accolades, including:

  • The Boston Globe's 2023 and 2024 Tech Power Players 50

  • Forbes' Top 50 Women in Tech

  • Fortune's 40 Under 40

After serving as Smart Eye’s Deputy CEO for a few years, el Kaliouby recently founded her own venture fund this year called Blue Tulip Ventures where she says the focus is to invest in “AI that’s good for people, good for the planet, is ethical, and is built with diverse teams.”

One of her recent investments was Etched, an AI chip startup that has secured $120 million in funding for their specialized chip designed for generative AI models, which is making waves for being a contender to Nvidia.

Follow her on Linkedin or Twitter. She’s the real deal. 😎

ENTERPRISE AI CASE STUDY

Toys "R" Us Sparks Debate with AI-Made Ad

Industry: Retail

Screenshot from Toys “R” Us advertisement

Toys "R" Us has made waves in the advertising world with its groundbreaking AI-generated commercial, created using OpenAI's text-to-video tool Sora.

This innovative approach marks one of the first instances of a major brand utilizing generative AI for a full-length advertisement, sparking both excitement and controversy in the industry.

The Ad's Content and Creation

The 66-second spot, titled "The Origin of Toys 'R' Us," depicts the childhood dreams of the company's late founder, Charles Lazarus, envisioning a magical toy store featuring the iconic mascot Geoffrey the Giraffe.

The ad was produced in collaboration with creative agency Native Foreign, using an early-access version of Sora, which is not yet available to the public.

The Wall Street Journal highlighted key aspects of the ad's creation, including:

  • It was produced in just a few weeks, condensing hundreds of iterative shots down to a couple dozen.

  • 80% of the work was done by AI, and then the agency then stepped in to make minor adjustments, similar to any postproduction process.

  • A team of about 20 individuals, including designers and animators, worked closely with the AI tool to refine the output.

Reception and Controversy

The ad has received mixed reactions from viewers and industry professionals. Many critics said it “sucks”, is “terrible and bad and awful” and amounts to “brand murder.”

Frankly, while the ad might be mediocre (and it's okay to say so), it's a bold move that deserves recognition. AI in advertising is here to stay, and this is just a taste of what's to come.

Hats off to Kim Miller Olko, chief marketing officer at Toys "R" Us. She stepped outside the box and embraced innovation with AI.

PRO TIP

How to Define a Winning AI Use Case

Not all AI ideas are created equal. Some are vague pipe dreams, while others are well-articulated plans ready for action. Here's how to refine your concept and turn it into a high-value AI use case.

  1. The Vague Notion:

    • Example: "Let's use AI to improve customer service."

    • Problem: Lacks specifics, doesn't define what "improve" means or how AI would be applied.

  2. Focused: Narrowing Down the Scope:

    • Example: "Let's use AI to enhance the customer support experience in our call center."

    • Improvement: This specifies the customer service area (call center) as the focus.

  3. Clear: Defining the AI Solution and Desired Outcome:

    • Example: "Implement an AI-powered chatbot in our call center to handle routine customer inquiries, reduce call wait times by 20%, and improve customer satisfaction ratings by 10% within the next quarter."

    • Why It Works:

      • Specific Problem: Long call wait times and potentially low customer satisfaction in the call center.

      • Clear AI Solution: AI-powered chatbot to automate routine inquiries.

      • Measurable Goals: 20% reduction in wait times, 10% improvement in satisfaction ratings.

      • Timeframe: Next quarter.

 😎 As my PhD supervisor used to exclaim when I would write general and vague statements, “specificity, Ayesha, specificity!”

See you next week!

-- The future awaits. Ayesha ♥️

I'd love your feedback...

Please vote below to help me improve the newsletter.

Login or Subscribe to participate in polls.