Home AdExchanger Talks From Hype To Hyperscale In AI

From Hype To Hyperscale In AI

SHARE:
Ikkjin Ahn, CEO & co-founder, Moloco

Nearly every ad tech company is out there promising some form of AI-powered magic. The vendor landscape is turning into a sea of sameness.

So what’s worthwhile and what’s worth chucking in the bin?

The best way to separate AI hype from reality is to roll up your sleeves and try out the tech for yourself, says Ikkjin Ahn, CEO and co-founder of machine learning-based ad tech startup Moloco, on this week’s episode of AdExchanger Talks.

It’s like watching a movie, he says. How do you know if it’s good before you even try it?

But there are certain elements that separate serious contenders from the rest of the pack, he says, and one of the biggest is the ability to perform at massive scale.

You can build “toy examples” very easily with generative AI, Ahn says, but “the key differentiator is how much you can feed the volume, how much scale you can achieve.”

That’s where AI hyperscale comes in, which involves running AI on huge, cloud-style infrastructure. It’s a mix of ultra-fast hardware and optimized software that’s been trained to power AI models faster and more efficiently than regular computers.

Moloco uses AI Hypercomputer, which is Google’s branded end-to-end supercomputing stack. It runs on Google Cloud and allows Moloco to process billions of requests a day at 10x speed and relatively low cost.

Online advertising is evolving from a reliance on carefully engineered features and audience segments to harnessing the power of these “foundational models” that can learn from huge volumes of raw data, Ahn says.

It’s a process, though.

While generative AI models are currently pretty good at answering questions, most are way too slow and expensive to meet the needs of real-time advertising, he says, which demands more powerful models that can operate instantly and affordably.

“Think about how long it takes to get an answer from ChatGPT,” Ahn says. “If your recommendation or sponsored search ad is taking three seconds, I mean, that’s a problem. And if any ad tech companies say, ‘Hey, I’ll burn like, what, $3 billion per quarter – that is not an answer, right?”

Also in this episode: How Moloco started its DSP business using Google’s display ad network as a test bed, how AI can help democratize commerce media beyond giants like Amazon and the role of creative automation in streamlining digital ad production.

For more articles featuring Ikkjin Ahn, click here.

Must Read

Pirated Sports Streams Are Warping TV’s Most Important Ratings

Although tides of ad revenue flow based on the ratings of certain tentpole TV events, a new crop of scammers now operate illicit sports livestreaming rings, and there’s almost nothing broadcasters can do about it.

AI Is Redefining Premium Content – Which May Not Be A Good Thing

At AdExchanger’s Programmatic AI conference, media experts discussed how the rise of AI-generated content is changing the industry’s understanding of “premium” content.

The Big Story Podcast

Prog AI Live: AI’s Slippery Slop

Recorded live in Las Vegas at Prog AI, the AdExchanger team tackles a tricky question: As AI floods the feed with chaotic, addictive content and people engage with it, what does “premium” even mean anymore?

Privacy! Commerce! Connected TV! Read all about it. Subscribe to AdExchanger Newsletters

The Programmatic Auction Is Changing In Real Time – Here’s How

Two decades after the first RTB auction, programmatic is more complex than ever – and that’s before you even consider generative AI.

Publicis Acquires LiveRamp In A Major Shakeup For Indie Data Collaboration

Hundreds of exasperated and unexpected ad industry phone calls were made on Sunday, as agencies and ad tech vendors discussed the fallout of Publicis Groupe’s $2.2 billion acquisition of LiveRamp over the weekend.

Finger connecting dots on a cork board network concept

These AI Agents Want To Handle All The Annoying Parts Of Media Buying

Meet Kovva, a new AI ad tech startup tackling the unglamorous gruntwork that programmatic has never fully automated.