Is AI Hype Already Dying!?

Here is what’s new in the AI world.

Gen AI’s Hype Will Be Dead Soon – Yann LeCun, Meta’s AI Chief

At the World Economic Forum in Davos, Yann LeCun, Meta’s AI Chief, dropped some big news. He thinks the AI we’re all hyped about right now like Generative AI and those fancy large language models, isn’t all that great.

Sure, they’re cool for basic stuff. But beyond that? Meh. And here’s the kicker: in three to five years, no one will even want to use them. He says, they’re basically on borrowed time.

So, what’s next?

Well, LeCun thinks there’s a whole new AI revolution coming. A big one. He’s talking about smarter systems, stuff that actually understands the world.

Current AI? It’s clueless.

It can’t remember much, doesn’t think ahead, and can’t plan worth a darn. Plus, it has zero common sense. Imagine a toddler with a fancy vocabulary but no idea what’s happening around them. That’s today’s AI.

But LeCun’s not just complaining. Meta’s cooking up something new in their labs. He says they’re trying to build systems that can, like, figure out how the world works just by watching and interacting with it.

If they nail this, we’re talking about AI with some real brainpower. Something that can think for itself. He thinks this could happen in the next five years. Maybe sooner!

He’s also saying robots are about to have their moment. The “decade of robotics,” he calls it. Is AI teaming up with robots? That could change everything. Think sci-fi, but, like, real life.

The stuff we use today might soon feel like an old flip phone. 

The future? 

Smarter, savvier, and a lot more human.

“Most Enterprises Are Thriving On AI” – Deloitte

Generative AI is a hot topic everywhere, right? Enterprises big and small are trying to figure out how to use it. But the good news? Most of them are actually getting it to work. And they’re seeing results.

Deloitte just dropped a report—’State of Generative AI Q4.’

They surveyed 2,773 leaders across 14 countries. The big takeaway? Most businesses are meeting or even crushing their expectations for ROI on Gen AI. Yep, even with the challenges.

A year ago, things were shaky. Companies were worried about data issues, and compliance.

But now- They’ve made progress. Big progress. Three-quarters of the respondents say their advanced Gen AI projects are hitting or exceeding ROI goals. Cybersecurity and IT are killing it, leading the charge.

Of course, it’s not all smooth sailing. On average, it’s taking organizations over a year to tackle adoption hurdles. Regulatory compliance? That’s the new top blocker.

And get this - only 40% of employees in most companies even have access to AI tools. Crazy, right? Jim Rowan from Deloitte says there’s a need for a cultural shift. Employees need tools and the knowledge to use them. 

Still, AI is saving time. It’s making people more efficient, especially in IT and customer service. 

Rowan calls out the software development lifecycle as a prime example - AI is helping from start to finish, from gathering requirements to testing. And contact centers? 

They’re getting a serious boost, automating tasks and speeding up service.

But here’s the kicker - Enterprises don’t move fast. Not like AI does. 

Rowan says businesses are “moving at enterprise speed.” 

Translation - Slow. They’re cautious, focusing on stability, governance, and getting the tech just right before scaling it up. It makes sense, but still.

What’s next? Agentic AI. That’s the big thing on the horizon. 

The report says 52% of companies are looking into AI agents. Multi-agent systems too. Rowan thinks this tech could transform how enterprises work - but only if they stop looking at individual use cases and start redesigning entire processes with AI at the core. 

Bold idea, huh?

How Will Gen AI Transform the Future of Agencies

Forty years ago, the ad industry looked a whole lot different. Back then, people were doing layouts by hand. Then came desktop publishing (DTP). Suddenly, it was all on computers. Fast forward, boom , the internet changed everything. 

SEO, e-commerce, social media, influencer marketing. 

Now? We’re staring down the barrel of another big change: generative AI. And yeah, everyone’s talking about what it means for creativity. But that’s just surface level. The real shake-up? It’s gonna hit the agency business model.

Creativity VS Commercial Creativity

Let’s get one thing straight. Creativity isn’t everything. I know, shocking, right? But here’s the deal, clients don’t pay for creativity. They’re buying what it does for their business. 

It’s all about commercial creativity. The creative idea is the drill - the business impact is the hole. The industry’s been too focused on the drill. Clients? They want results.

Look at Accenture or Deloitte. Or Meta and Google, cutting out agencies entirely. Agencies are grabbing smaller and smaller pieces of growing marketing budgets. Not a good look!

The business model breakdown

Here’s where Gen AI flips the script. Agencies juggle two kinds of work:

1. Big-idea work (BIW): the flashy stuff - TVCs, campaigns. Makes agencies proud but rarely profitable.

2. Business-as-usual (BAU): the day-to-day grind - banners, social posts, website updates. It's not ‘sexy' per say, but predictable and profitable.

Agencies balance these two to stay afloat. But gen AI? It’s coming for BAU work. Why pay an agency when AI can whip up social posts for a fraction of the cost? Clients are catching on.

Imagine this -

• An agency gets 70% of its revenue from BAU and 30% from BIW.

• BAU makes a 20% profit; BIW loses 20%. Overall, they’re making an 8% profit.

• Now, the client takes half of BAU in-house. The agency’s revenue drops. Their profit shrinks to just 1.6%. Ouch.

To survive, agencies would need to either streamline big-idea work (hard) or charge clients more (even harder).

Agencies have two choices going forward

1. Go upstream-

   • Become a creative marketing consultancy. Solve real business problems with creative, strategic thinking.

   • Charge based on value, not hours. High risk, high reward.

2. Go downstream- 

   • Focus on what AI can’t do - yet. Offer creative toolkits for clients to run with.

   • Sell ideas, not execution. Standardized processes = high margins.

Both paths have challenges. Going upstream means convincing clients you’re worth it. Going downstream means relying on clients to handle AI-powered execution themselves.

But one thing’s clear - the old model is running out of time. Agencies need to evolve - or risk getting left behind.

OpenAI Unveils Operator – Your Autonomous AI Assistant

Sam Altman, OpenAI’s CEO, kicked off the year with a bold prediction: 2025 will be the year of AI agents. These tools? They’ll automate tasks and make decisions for you. And now, OpenAI’s taking its first real shot with Operator.

What’s OpenAI’s Operator?

It’s a new AI agent. Think of it as your online assistant. It can browse the web, book your flights, make dinner reservations, and even shop for you. OpenAI announced a research preview last Thursday. For now, it’s only available in the U.S., and you’ll need the $200 ChatGPT Pro plan to try it.

Other plans? They’ll get it later.

Europe? Gonna have to wait.

Altman said it himself during a livestream.

You can find Operator at operator.chatgpt.com. Eventually, OpenAI wants it integrated into all ChatGPT apps.

How does this ‘Operator’ work?

When you turn on the Operator, a little window pops up. Inside it is a dedicated browser where it works its magic. It navigates websites like a human. Clicking buttons. Filling out forms. Picking from drop-down menus. And yes, you can take control at any time.

Operator runs on OpenAI’s Computer-Using Agent (CUA) model. This combines visual smarts from GPT-4o with advanced reasoning from other models. Unlike traditional tools, it doesn’t rely on APIs. It works directly with website interfaces. Pretty cool, huh?

OpenAI’s even teaming up with big brands like DoorDash, Uber, Instacart, and StubHub to ensure Operator respect their rules. No shady business here.

OpenAI promises that the Operator doesn’t store or screenshot your data. On high-risk sites, like email, you’ll have to supervise it directly. The Operator is not flawless. But it’s a solid start. 

Altman’s 2025 prediction has started to look real. The future of AI agents is closer than we think.

Elon Musk stirred the pot too. 🍯

He claims OpenAI, Oracle, and Softbank don’t have the cash to back their massive $500 billion AI project. This big venture, dubbed the Stargate project, was just unveiled by Donald Trump earlier that day at the White House.

“They don’t actually have the money,” Musk posted on X, replying to OpenAI’s announcement about Stargate. He didn’t stop there. “Softbank has well under $10B secured. I have that on good authority,” he added.

Musk’s comments landed just hours after Trump’s bold claim - Stargate would invest $500 billion and create over 100,000 American jobs ‘almost immediately.’ Trump called it “something very special.”

And as we all know Musk, he did not buy it!

/It’s Musk VS Trump - Stargate AI Investment Heats Up

Elon Musk stirred up trouble Tuesday night.  He claims that OpenAI, Oracle, and SoftBank don’t have the cash to back their massive $500 billion AI project, Stargate. This came just hours after Donald Trump announced the initiative at the White House.

“They don’t actually have the money,” Musk wrote on X, replying to OpenAI’s post about Stargate. He doubled down, saying, “SoftBank has well under $10B secured. I have that on good authority.”

Sam Altman, OpenAI’s CEO, wasn’t having it. By Wednesday morning, he hit back directly at Musk.

Then he threw some shade. “I hope you’ll mostly put 🇺🇸 first,” he added.

SoftBank? No comment. OpenAI and Oracle? Silent too. But a source close to the project called Musk’s remarks ‘far off base.’ They hinted the bad blood between Musk and Altman might’ve fuelled the drama.

Trump’s big bet

During the announcement, Trump called Stargate ‘very special.’ He promised it would invest at least $500 billion and create “over 100,000 American jobs almost immediately.”

Trump wasn’t alone. Sam Altman, Oracle’s Larry Ellison, and SoftBank’s Masayoshi Son stood beside him. Ellison even announced that 10 massive data centers are already under construction in Texas, with 10 more on the way.

The Microsoft angle

Microsoft is also on Stargate. OpenAI listed them as a key tech partner, and Satya Nadella, Microsoft’s CEO, said the company plans to pump $80 billion into AI-capable data centers next fiscal year.

When CNBC’s Squawk Box asked Nadella about Musk’s funding claims, he shrugged it off. “All I know is, I’m good for my $80 billion,” he said.

The drama runs deep but 

Musk’s rocky history with Altman adds more spice to the story. In March, Musk’s AI startup xAI sued OpenAI, accusing them of breaching a contract. He later dropped the suit but sued Altman directly in August.

Despite the tension, Altman’s response stayed cool—publicly, at least. Back in December, he said, “I strongly believe Elon will do the right thing.”

For now, Stargate is moving forward. OpenAI says $100 billion will be deployed immediately, with Son serving as chairman. Oracle shares even jumped 7% after the announcement. But Musk’s doubts linger.

And the AI battle? It’s just heating up.

Your opinion matters!

Hope you loved reading our piece of newsletter as much as we had fun writing it. 

Share your experience and feedback with us below ‘cause we take your critique very critically. 

Thank you for reading

-Shen & Towards AGI team