The AI Economy: AI at CES 2025 and OpenAI's New o3 Model

This is "The AI Economy," a weekly LinkedIn-first newsletter about AI's influence on business, work, society and tech and written by Ken Yeung. Sign up here.

IN THIS ISSUE: Welcome to the final newsletter of 2024. Thank you for subscribing to and reading “The AI Economy.” Next month marks one year since I launched this publication, and I appreciate the reception it has received. In the coming year, expect more stories about AI’s impact on business, work, technology, and society.

Regarding today’s edition, we explore what’s in store for this year’s Consumer Electronics Show and how AI will play a central role. And surprise! OpenAI announces a new frontier reasoning model.

Get Ready for an AI CES

We’re a couple of weeks away from a new year and the unveiling of the latest gadgets and devices we never thought we’d need at home or in our lives. That’s right: The annual Consumer Electronics Show is right around the corner, and you can bet much of what’s there will be heavily influenced by artificial intelligence.

This isn’t the first CES to showcase AI, but with the rapid pace of AI advancements in 2024, the technology will play an even more central role in the newest products at the 2025 show. Whether it’s robotics, autonomous vehicles, surveillance equipment, computers, or home entertainment, come January 6, Las Vegas will all be about “AI, AI, AI!”

Underscoring the prominence organizers place on the technology, Nvidia Chief Executive Jensen Huang will give a keynote address to kick off CES. With the company playing a central role in AI’s development, all eyes will be on him as he outlines the future of global industries, thanks to AI and accelerated computing. What will the next twelve months look like for the technology we use?

At CES 2025, I’m looking for a vision of how tech makers bring AI to consumers and businesses in more relatable ways. It’s time to see what AI-powered products for everyday use will look like—not just concept devices but those genuinely “coming soon to the market.”

Over the past year, we’ve seen the rise of AI-powered wearables, like ones from Rabbit and Humane, though with disappointing results. Expect more such hardware to be showcased at CES, built by startups looking for their breakout moment.

Robots will also take the spotlight, with most designed for industrial, manufacturing, healthcare, retail, and airport applications. However, there could be a few geared toward personal use as well.

And let’s not forget about autonomous vehicles being at CES. The show floor will be filled with companies showcasing technology powering self-driving cars, mapping the real world, and other intelligent applications.

One thing I’ll be watching closely is the diversity of AI companies at the show. A glance at the exhibitor directory suggests that many of the companies appearing at the Las Vegas Convention Center will hail from Asia, particularly China, Japan, and Korea. Could this be an indicator of where innovation is coming from?

As for my prediction about the “winner” at CES 2025, I’d wager that it’ll be an AI wearable.

What’s your bet?

Featured Image: The Consumer Electronics Show logo hanging in the Venetian resort in Las Vegas, Nevada, in 2020. Photo credit: Ken Yeung

OpenAI Releases o3 Model

OpenAI’s Hongyu Ren, Mark Chen, and Sam Altman discuss o3. Image credit: YouTube Screenshot

For the grand finale of its “12 Days of OpenAI,” OpenAI unveiled o3, a new reasoning model. It’s a successor to OpenAI’s o1 model and comes in two variations, o3, and o3-mini, a smaller model suited for specialized tasks and is described as good at performance and cost.

“We started this 12-day event 12 days ago with the launch of 01, our first reasoning model. It’s been amazing to see what people are doing with that and very gratifying to hear how much people like it,” Chief Executive Sam Altman said during his opening remarks. “We view this as sort of the beginning of the next phase of AI where you can use these models to do increasingly complex tasks that require a lot of reasoning.”

And no, you didn’t miss anything—there was no “o2.” Altman explained that OpenAI chose o3 for its new frontier model name “out of respect to our friends at Telefonica” and also “in the grand tradition of OpenAI being really, truly bad at names.”

o3 is the final chapter in a series of announcements OpenAI has unveiled over the past couple of weeks. Highlights include the launch of ChatGPT Pro, the text-to-video tool Sora, ChatGPT’s integration with Apple Intelligence, expanded ChatGPT Search access, and more. The company is making its newest model available to safety researchers and plans to make o3-mini publicly available in January, with o3 launching afterward.

OpenAI says o3 achieved an 87.5 percent score on ARC-AGI, a test that evaluates how efficiently AI systems can acquire new skills outside the data they were initially trained on. By comparison, OpenAI’s o1 model scored between 25 percent and 32 percent.

“We believe [o3] represents a significant breakthrough in getting AI to adapt to novel tasks,” writes Francois Chollet, the creator of ARC-AGI, on social media. He elaborated further in a blog post, stating that this is a “surprising and important step-function increase in AI capabilities, showing novel task adaptation ability never seen before in the GPT-family models. For context, ARC-AGI-1 took [four] years to go from 0 percent with GPT-3 in 2020 to 5 percent in 2024 with GPT-4o. All intuition about AI capabilities will need to get updated for o3.”

Here’s how it performs against other benchmarks, according to TechCrunch:

The model outperforms o1 by 22.8 percentage points on SWE-Bench Verified, a benchmark focused on programming tasks, and achieves a Codeforces rating — another measure of coding skills — of 2727. (A rating of 2400 places an engineer at the 99.2nd percentile.) o3 scores 96.7% on the 2024 American Invitational Mathematics Exam, missing just one question, and achieves 87.7% on GPQA Diamond, a set of graduate-level biology, physics, and chemistry questions. Finally, o3 sets a new record on EpochAI’s Frontier Math benchmark, solving 25.2% of problems; no other model exceeds 2%.

▶️ Watch OpenAI’s o3 announcement on YouTube

Microsoft Claps Back at Salesforce as AI Rivalry Keeps Going

Salesforce CEO Marc Benioff can’t help but boast about how his Agentforce platform outshines Microsoft’s Copilot. He’s especially proud that the public and media have taken notice of his labeling Copilot as “Clippy 2.0.” On Tuesday, Benioff again went after the Windows maker, claiming customers can’t find any mention of Copilot on Microsoft’s website.

Microsoft may have had enough. During Benioff’s event, the company published its 2024 year in review, detailing notable use cases of its AI. It countered Salesforce’s claim that Microsoft customers weren’t finding any value in using Copilot.

On Friday, Charles Lamanna, Microsoft’s Corporate Vice President for Business and Industry Copilot, took to LinkedIn to address an allegation made by Benioff that Microsoft wasn’t using Copilot internally. While praising the Salesforce CEO’s passion, Lamanna called his accusation “not true.”

“I use Copilot. My team uses Copilot. And pretty much all of Microsoft has been actively using Copilot for two years. That’s 100s of thousands of us!” he writes. The executive goes on to highlight several use cases, such as an agent in Azure.com that boosted engagement 4.75 times and drove a 21.5 percent increase in conversion rates; how Microsoft’s customer support, HR, and legal teams used Copilot to resolve cases faster; and boosted revenue per seller by 9.4 percent and increased close rates by 20 percent.

Naturally, this won’t end the AI rivalry between Salesforce and Microsoft, although Benioff seems to be the one instigating things and perhaps punching up-market. Nevertheless, expect more claims like this to surface over the coming year.

Today’s Visual Snapshot

One definition of an AI agent. Image credit: Jeremiah Owyang/Blitzscaling Ventures

What is an AI agent? Everyone may be talking about the same thing, but some people have differing opinions about what defines these sophisticated chatbots. Venture capitalist and analyst Jeremiah Owyang recently presented the above slide to Nvidia with one such perspective. This is what it means when he says “AI agent,” their attributes, and examples.

Quote This

“We can see that to unlock this GDP growth, you need a breakthrough technology. We have to become a digital labor provider. So, while we, for 25 years, have been helping companies manage and share their information [and] store their data, now, here we are delivering digital labor. And digital labor is this new horizon for business—this idea that a door has opened for business, and that business will never be the same. [The way] we architect our businesses and run our businesses and staff our businesses, and think about our businesses will never be the same.”

— Salesforce Chief Executive Marc Benioff discusses the significant impact that AI agents have on how businesses are built and how they will transform the workforce.

This Week’s AI News

🔬 Science and Breakthroughs

💼 Business, Marketing, Media, and Consumer Applications

ChatGPT comes to WhatsApp—here’s how to message the AI chatbot for free (Tom’s Guide)
How OpenAI hopes to sever its nonprofit roots (The New York Times)
How LinkedIn is using AI-powered ad tools to lure B2B marketers (Digiday)
Boston Dynamics lays off employees, says it’s “burning through cash” (Futurism)

⚖️ Legal, Regulatory, and Ethical Issues

Every AI copyright lawsuit in the U.S., visualized (Wired)
Congress publishes long-awaited bipartisan AI report, leaving some tech-watchers on edge (Politico)
UK arts and media reject plan to let AI firms use copyrighted material (The Guardian)
Philadelphia wants to be a national AI in education model (The 74)
Educators: Without AI literacy, students will be “unprepared for the future” (EdWeek)
The artist making unsettling AI images of the human body (CNN)

💥 Disruption, Misinformation, and Risks

Can AI automate the writing of review articles? (IEEE Spectrum)

End Output

Thanks for reading. Be sure to subscribe so you don’t miss any future issues of this newsletter.

Did you miss any AI articles this week? Fret not; I’m curating the big stories in my Flipboard Magazine, “The AI Economy.”

Follow my Flipboard Magazine for all the latest AI news I curate for “The AI Economy” newsletter.

Connect with me on LinkedIn and check out my blog to read more insights and thoughts on business and technology.

Do you have a story you think would be a great fit for “The AI Economy”? Awesome! Shoot me a message – I’m all ears!

Until next time, stay curious!

Subscribe to “The AI Economy”

New issues published on Fridays, exclusively on LinkedIn

CES 2025: Now With More AI

Get Ready for an AI CES

OpenAI Releases o3 Model

Microsoft Claps Back at Salesforce as AI Rivalry Keeps Going

Today’s Visual Snapshot

Related Reading:

Quote This

This Week’s AI News

🏭 AI Trends and Industry Impact

🤖 AI Models and Technologies

✏️ Generative AI and Content Creation

💰 Funding and Investments

☁️ Enterprise AI Solutions