Today in AI

• News roundup

• Today’s big story

• This week on the podcast

• AI events

News roundup

The top stories in AI today.

NEW LAUNCHES

The latest features, products & partnerships in AI

• Pinterest boosts ad campaigns with new AI and automation tools

• Perplexity just released 2 new features — here’s what’s inside

• Renowned horror film studio teams with Meta to leverage AI video tech

• ChatGPT Canvas adds tracked changes feature

• Adobe unveils AI-powered 3D rotation tool for 2D drawings

• An inside look at NotebookLM’s new AI-powered features

IMPLEMENTATION

Announcements, strategies & tools

• AI stagnation and the gap between investment and adoption

• AI adoption will require solving these massive LLM security vulnerabilities

• Why ‘predictive’ generative AI is the smart bet for real returns

• How AI is transforming the construction industry

• How Walmart is evolving its data analytics platform around AI

MODELS

Deployment, research, training & infrastructure

• Recent studies by OpenAI and Apple challenge AI model progress

• Apple internally believes it is 2 years behind competitors in AI development

• 5 AI predictions Lex Fridman got right for 2024

• Meta just released Spirit LM, an open-source multimodal AI model

• Meta shares massive AI materials database to accelerate research

• How AI can help find common ground in group deliberations

MONEY FLOWS

Economics, Wall Street, investing & fundraising

• Microsoft and OpenAI consider the value of $14B in for-profit switch

• China’s Pony.ai files for US IPO to advance autonomous driving

• The economics and ROI of big tech’s AI spending

• Terray Therapeutics secures $120M for AI-driven drug discovery

What’s happening in AI right now

Tech giants are doubling down on models

The AI landscape is rapidly evolving, with tech giants pouring unprecedented resources into development. However, not all industry leaders are on equal footing in this high-stakes race. Recent reports suggest that Apple, long considered a pioneer in consumer technology, is struggling to keep pace with its competitors in the model arena.

The Cupertino conundrum

Internal assessments at Apple indicate the company is lagging approximately two years behind its rivals in AI development. This revelation comes as a surprise to many, given Apple's reputation for innovation and its vast resources. The company's virtual assistant, Siri, once a groundbreaking technology, now underperforms compared to more advanced AI models like ChatGPT.

Apple's strategy to close this gap involves a two-pronged approach: developing on-device AI models and forging external partnerships, including a notable collaboration with OpenAI. This hybrid method aims to leverage Apple's strengths in hardware integration while integrating cutting-edge AI capabilities developed by specialists.

The company has set an ambitious goal to integrate AI across its entire product ecosystem by 2026. This timeline underscores the urgency Apple feels to catch up in the AI race, recognizing that falling behind could have significant long-term consequences for its market position.

Big tech's billion-dollar bets

While Apple plays catch-up, other tech giants are making massive investments in AI development. Google has announced plans to invest over $100 billion in AI over time, while Microsoft has entered a $10 billion power purchase agreement for AI infrastructure. These staggering figures reflect the industry's belief that early leadership in AI will translate to long-term market advantages.

Nvidia, a key player in the AI chip market, reported that its top customer spent $4.2 billion on chips and services in a single quarter. This level of spending raises questions about the economics and return on investment of AI development. As one industry observer noted, "The true value of these investments will depend on practical applications and revenue generation" in the long term.

Multimodal models and vertical applications

As the AI arms race intensifies, the focus is shifting towards multimodal models that can process various types of data, including text, speech, and images. Meta's recent introduction of Spirit LM, an open-source multimodal language model, exemplifies this trend. Spirit LM integrates text and speech inputs and outputs, aiming to provide more expressive and natural-sounding speech generation.

These advancements in multimodal AI are unlocking new opportunities for vertical applications across industries. Improved voice and vision capabilities are transforming workflows in sectors ranging from healthcare to manufacturing. At the same time, AI agents are showing promise in complex tasks such as sales, negotiations, and investigations.

The Habermas Machine

In a fascinating development, Google DeepMind researchers have created an AI system dubbed the "Habermas machine," designed to mediate group discussions on complex social and political issues. This system uses two large language models to generate and evaluate statements reflecting group opinions and areas of agreement.

Initial studies show promising results, with participants preferring AI-generated statements over human mediators 56% of the time. The system demonstrated the potential to decrease division on issues after deliberation. However, challenges remain, including the lack of robust fact-checking capabilities and ethical considerations surrounding AI's role in shaping public discourse.

Looking ahead

The rapid advancement of AI technologies, particularly in models, is set to fundamentally change work processes and interactions across industries. Novel business models, including AI copilots and AI-enabled services, are likely to emerge as these technologies mature. What steps should companies take to stay competitive in this rapidly evolving model landscape?

We publish daily research, playbooks, and deep industry data breakdowns. Learn More Here

This week on the podcast

Can’t get enough of our newsletter? Check out our podcast Future-Proof.

In this episode, the hosts Anthony Batt and Shane Robinson with guest Joe Veroneau from Conveyor discuss outsmarting paperwork. Conveyor is a company that helps automate security reviews and document sharing between companies. They use AI technology, specifically language models, to automate the process of filling out security questionnaires. This saves customers a significant amount of time and improves the quality of their responses.

AI events

Best way to get AI literate? Go to some awesome events.

We’re thrilled to share that COAI is partnering with HumanX 2025—the AI conference that’s set to redefine the future of technology. Taking place on March 10-13, 2025 at The Fontainebleau Las Vegas, this forum is where the brightest minds in AI will gather to shape what’s next. And we want you to join us!

Why attend HumanX?
HumanX isn’t just another tech conference. It’s a unique opportunity to:

• Connect with industry leaders, C-suite executives, policymakers, and innovators from around the globe.

• Learn from top-tier speakers like Kevin Weil, Clara Shih, and Sridhar Ramaswamy about the latest AI trends and how they’re transforming cross-functional industries.

• Explore personalized strategies and solutions to drive your business forward with AI.

Whether you’re a startup, an established business, or an AI pro looking to make meaningful connections, HumanX is the place where AI meets opportunity.

Exclusive offer for our community
As a valued member of our community, we’re excited to extend an exclusive offer to attend HumanX 2025. Register now with our code HX25p_coai and save $250 on general admission!