NEWS YOU CAN’T MISS

Is Your Next Friend an AI?

Platforms like Character.ai are leading the charge, boasting 250M+ monthly visits and 20M monthly users. Users spend an average of 2 hours daily chatting with AI personas ranging from celebrity impersonators to therapists.

Seeing Character.ai’s success, Meta and Google are jumping in (Google is building its own AI character platform). They want more user engagement and they realize that people are totally fine with having AI friends.

ps. Character.ai now offers voice calls with its AI agents. For example, you can call your AI friend Émilie from Paris and practice your French with her. C'est magnifique, non? 🇫🇷

You Have to Try Claude 3.5

Anthropic has released Claude 3.5 Sonnet and it's mind-blowing. Twitter is abuzz with everyone singing its praises. I’ll give you just one example to try.

When I started learning to code, one of the first assignments I had to do was build a Tic-Tac-Toe game. I asked Claude to “create an interactive Tic-Tac-Toe game using React,” and it not only wrote all the code but also created the app itself for me (see picture above).

Yes, Claude is fast, personable, and smart, and leading in multiple categories of the LMSYS Chatbot Arena, a reputed platform for crowdsourced evaluation of large language models.

But competitors like ChatGPT will likely catch up soon. What I find really interesting about Claude 3.5 is its product innovation.

Unlike other chatbots (think Open AI’s ChatGPT, Google’s Gemini) where all conversation happens in one chat, Anthropic has created what it calls an Artifact (see the document on the right hand side in the picture). This means it shows me what it’s generating for me in one document, which I then can edit and add to as I like.

This is the first Generative AI interface innovation we’ve seen in the market and it’s super useful and user-friendly.

Anthropic states that Artifacts will soon be expanded with new features, including tools for collaboration with larger teams and knowledge base storage.

💫Instagram co-founder Mike Krieger's recent appointment as Anthropic's Chief Product Officer likely explains these recent innovations.

What we are seeing is the evolution of large language models from just chatbots to products.

Check it out!

TREND ALERT

Defense Tech and AI: A New Era of Warfare

Defense tech startup Anduril nears $1.5B funding at a $12.5B valuation after doubling revenue to $500 million.

Anduril, a growing supplier to the US armed forces, develops autonomous weapons and defense systems including drones, rocket launchers, and submarines. For example, it recently secured a $19 million U.S. Navy contract to develop solid rocket motors for the Standard Missile-6.

With increasing conflicts and geopolitical tensions in the world, the rise in AI-powered defense startups is a trend you need to know.

For example, San Diego-based Shied AI is building AI pilots for military aircraft and drones. Shield AI and the Pentagon have collaborated on the Air Combat Evolution (ACE) project, using an AI-powered F-16 fighter jet in advanced dogfighting scenarios (see picture above of the jet). The tests included high-speed, close-range engagements to push the boundaries of autonomous flight.

Another notable one is Munich-based Helsing, which leverages AI to analyze massive amounts of data from sensors and weapons systems, providing real-time battlefield intelligence and aiding military decision-making. Helsing's software is being utilized to enhance the AI capabilities of drones in Ukraine, and it’s in talks to raise another $500m, tripling its valuation to $4.5B.

Anduril's founder, 31-year-old Palmer Luckey, explained to the Financial Times how he discusses defense tech and AI with investors and military clients: “Xi Jinping thinks he can out-innovate American defense. Is he right?“

WOMEN IN AI

Meet Joy Buolamwini

Celebrating this week's woman in AI 🥳: Meet Dr. Joy Buolamwini, founder of the Algorithmic Justice League (AJL) and a digital activist.

Born in Canada, Buolamwini holds a master's degrees from Oxford University (as a Rhodes Scholar) and a PhD in Philosophy from MIT Media Lab.

Her MIT thesis, "Gender Shades," demonstrated the significant disparities in accuracy rates for facial analysis software when identifying darker-skinned individuals, especially women. This groundbreaking research exposed the discriminatory potential of AI and sparked widespread discussion about algorithmic bias and the importance of diversity in tech.

In 2016, Buolamwini founded the Algorithmic Justice League (AJL) to raise awareness about the impacts of AI, equip advocates with empirical research, and promote more equitable and accountable AI systems.

She has received numerous accolades, including being named to TIME's 100 Next list in 2023, Fortune's World's 50 Greatest Leaders in 2019, and MIT Technology Review's 35 under 35 in 2018.

I’m a big fan!

ENTERPRISE AI CASE STUDY

Inside Goldman Sachs' Generative AI Strategy

Industry: Financial Services

The financial giant Goldman Sachs is rolling out its first generative AI tool company-wide.

According to the Wall Street Journal, by the end of June, thousands of programmers across the firm will have access to an AI-powered coding assistant, promising a remarkable 20% increase in efficiency.

The generative AI tool, developed on the company's internal platform, integrates models from Microsoft, Google, and Meta. Goldman Sachs has explicitly banned the use of OpenAI’s ChatGPT at work.

As my friend highlighted to me, this approach is typical of Goldman Sachs, which has a history of not just taking platforms off the shelf, but fine-tuning them to their own specific needs.

Consistent also with its conservative approach, Goldman Sachs decided to keep tight control of its generative AI platform centrally. Contrast this with biotech firm Moderna that has an enterprise relationship with OpenAI where every team can build its own ChatGPT agent.

According to CIO Marco Argenti, the speed at which developers are building generative AI applications has shortened from months to weeks thanks to the coding assistant.

So they are working on launching a number of products like a proprietary research summarization tool that gives advisors insights.

Move over, Gordon Gekko, AI is the new star on Wall Street! 😎