NEWS YOU CAN’T MISS

Is AI the Future of Film? Ashton Kutcher Thinks So

Hollywood actor Ashton Kutcher commented on the rise of AI-driven video production. Kutcher, who has access to a beta version of Sora, OpenAI's generative video tool, described it as "pretty amazing." 

According to Kutcher, Sora can generate realistic 10-15-second videos, which leads him to predict that AI will soon have the capability to "render a whole movie." 

Leap of faith from 10 seconds to a movie? Not really. Kutcher says that since he saw Sora 6 months ago, it has made tremendous progress. Plus, new processors from Nvidia will make AI perform 30 times better, further making tools like Sora exponentially better.

The bottom line: Runway's new AI model received many positive comments for depicting complex scenes and human emotions. Google Deepmind also released an AI model that generates synchronized soundtracks for such videos. But instead of these individual stories, we need to pay attention to fact that is AI getting better … faster. 

😎 Kutcher got a lot of pushback on Twitter for his comments, but I agree with Kutcher that we can't put our heads in the sand like an ostrich. AI will massively disrupt movie production from Hollywood to Bollywood.

ps. I wouldn't take him lightly: Kutcher has made a ton of money spotting trends and investing in startups like Uber, Airbnb, Spotify, and Warby Parker.

Why Is Everyone Talking About Robots?

The buzz around humanoid robots has reached a fever pitch, with Silicon Valley VCs all talking about "physical AI" at every meetup and conference. 

At Tesla's shareholder meeting, Elon Musk said he already has a few Optimus robots in the California plant and predicts that "next year we'll have over a thousand, maybe a few thousand, Optimus robots working at Tesla." 

Interestingly, he said that buying a humanoid robot would cost around $20,000 once Tesla hits robot production at scale.

NVIDIA CEO Jensen Huang is equally optimistic, declaring that "the next wave of AI is physical AI," In Taiwan this month, Huang emphasized NVIDIA's commitment to developing advanced robotics platforms. 

The reason for all this enthusiasm is that robots are getting more intelligent thanks to AI in two ways:

• You can give robots tasks in natural language or upload a manual to read (pretty much how humans interact with AI agents like ChatGPT). 

• Companies like Nvidia are developing AI models that will enable robots to learn movements quickly by observing humans or videos and making decisions on how to interact with and move in the world around them.

If this sounds far-fetched, look at the automotive companies apart from Tesla starting pilots with robots:

• BMW: Announced plans to test humanoid robots from Figure AI at their South Carolina manufacturing facility.

• Mercedes-Benz: Exploring the use of Apptronik's Apollo humanoid robot for tasks like inspecting and delivering components to production line workers.

😎 It's still early days in humanoid manufacturing, but it's a trend we cannot ignore. 

Trend Alert: The Data Center Boom

Every single big corporation has AI on its roadmap, but AI needs more than just the goodwill of CEOs to succeed: it needs data centers, and data centers need power.

This is the behind-the-scenes AI infrastructure worth discussing.

First, data centers are popping up everywhere . About a third of the world's 8,000 data centers are in the US, but their expansion is a global trend. 

Here are two examples: Microsoft has committed $3 billion in Virginia and Indiana, $3.3 billion in Southeast Wisconsin, and $7.16 billion in northeastern Spain to build data centers. Amazon is investing $11 billion in Indiana and $9 billion in Singapore.

But how will we power these data centers? 

One emerging popular option is nuclear power, particularly small modular reactors (SMRs). Nuclear energy offers a stable and reliable source of electricity with low greenhouse gas emissions. 

Take Microsoft co-founder and philanthropist Bill Gates. Gates has invested $1 billion into a nuclear power plant in Kemmerer, Wyoming. The other investor is the US Department of Energy. Amazon also recently paid $650 million for a nuclear-powered data center in Pennsylvania.

Food for thought: Ami Badani, Chief Marketing Officer of semiconductor firm Arm, highlighted that data centers training AI models now use 2% of the world's electricity and cautioned that AI could account for 25% of US power consumption by 2030. 

🌎 Finding a sustainable way to cater to AI's insatiable need for data centers and power is critical to our planet's future.

WOMEN IN AI

Meet Wardah Inam, CEO of Overjet

Celebrating this week's woman in AI 🥳: Meet Wardah Inam, CEO and Co-founder of Overjet, a pioneer in AI for dentistry.

Wardah, born in Pakistan, completed her PhD in electrical engineering from MIT and launched Overjet in 2018.

Overjet’s platform, developed with AI trained on millions of dental X-rays, serves thousands of dentists across North America.

It received FDA clearance to detect, outline, and quantify oral diseases in X-rays. For example, dentists can identify 32% more tooth surfaces containing cavities with Overjet.

Wardah has been recognized for her work:

• Featured in Boston Globe's Tech Power Players 50, 2023

• Named to Inc. Magazine's Female Founders 100 list in 2022

• Invited to speak at Fortune's Most Powerful Women Next Generation summit

Despite the current "VC Winter" - with funding rounds decreasing and startups shuttering - Wardah raised Overjet’s Series C, valuing the company at $550 million.

Fast Company recognized Overjet in its "The 10 most innovative companies in Healthcare of 2023" list, while Forbes named it one of the "Top 50 Most Promising AI Companies Shaping the Future."

Poor oral health is now understood to be linked to broader health issues, including Alzheimer's, making Wardah’s work even more crucial.

😳 ps. According to Harvard Health, 90% of adults have had at least one cavity. I’m going to the dentist next week. What about you?

ENTERPRISE AI CASE STUDY

AI Chaos at McDonald‘s Drive-Thru

Industry: Fast Food

Many AI projects fail miserably. So today, let’s have a look at this week’s news on McDonald's AI fiasco at its drive thrus.

An AI-powered voice ordering technology was developed in partnership with IBM and launched at 100 McDonald’s drive-thru locations across the United States.

Key lesson: please test, test, test your AI systems before deploying them for customers. In today’s viral social media world, you can’t get away with bad customer experiences.

This will save you from a McFlurry of AI-related headaches down the line. 😉