The Drum Awards Festival - Extended Deadline

-d -h -min -sec

Deep Learning AI Digital Advertiisng

How artificial intelligence is transforming the world in 2023

RTB House


Open Mic article

This content is produced by a publishing partner of Open Mic.

Open Mic is the self-publishing platform for the marketing industry, allowing members to publish news, opinion and insights on

Find out more

April 25, 2023 | 7 min read

From HAL 9000 in 2001: A Space Odyssey to Ava in Ex Machina, artificial intelligence (AI) has, for years, taken centre-stage in the realm of science fiction

From HAL 9000 in 2001: A Space Odyssey to Ava in Ex Machina, artificial intelligence (AI) has, for years, taken centre-stage in the realm of science fiction. However, what was once the stuff of fantasy is fast-becoming a reality. In recent years, AI has experienced rapid growth and specifically with the development in deep learning technology it now means machines can process vast datasets, and, more importantly, learn from it to make accurate predictions that surpass any human capability. AI and deep learning are already transforming the world we live in, e.g., cancer screenings, displaying ads, and the recently launched ChatGPT, but the future possibilities of this technology are endless.

While ChatGPT is seriously cool, it’s just a small part of a much bigger picture.

What is AI?

Here’s what ChatGPT says AI is:

In other words, it’s a computerized system that can analyze huge amounts of data to find correlations and patterns in a much quicker time-frame than a human. And once patterns are found, they can be used to identify medical conditions, create a chatbot, or understand what a customer is most interested in purchasing.

It’s also important to note that when people talk about AI, they are actually referring to machine learning or deep learning algorithms.

What are the two main AI algorithms?

Machine learning and deep learning both involve training a computer system to recognize patterns in data. And while the terms are used interchangeably, there are significant differences between the two.

Machine learning encompasses a variety of techniques such as regression, decision trees, and support vector machines. It analyzes data to make decisions or predictions and learns and adapts with experience without specific programming. However, human input is required to correct errors and determine which features of the data are relevant for making predictions.

Deep learning, on the other hand, was built to correct the machine learning’s need for human intervention. Artificial neural networks, modelled on the structure of the human brain, allow data to pass through a web of interconnected algorithms and processed similarly to how humans process information. Deep learning learns to recognize patterns in data and improve its accuracy with repetition. As a result, deep learning is significantly more accurate than machine learning algorithms.

Why has AI suddenly gone mainstream?

While AI has been around for decades, it’s only in recent years it has gained popularity. But it was the launch of ChatGPT, a chatbot that mimics human speech, in December 2022 that largely helped take AI mainstream. The platform hit one million users in just five days, while it took Facebook 10 months.

ChatGPT aside, there are other reasons why AI has become popular in recent years:

  • Big data has made it possible to train deep learning models to provide more accurate and useful predictions
  • Advancements in computing power and cloud technology means it’s possible to process and analyze data at speed and scale, making AI more accessible and affordable to all businesses
  • Smart devices and the internet of things (IoT) has helped create the data used to train and improve AI models
  • A growing demand for personalized, on-demand services has fuelled AI adoption that provide users with personalized experiences and businesses with cost-effective solutions

Although ChatGPT, and other similar solutions are a huge leap forward, they are still not ‘true intelligence’. ChatGPT has access to the entirety of GitHub and a vast database to draw from to produce code or write articles and speeches, but is not connected to the internet. And while it can learn and produce convincing content, it’s not capable of replacing humans in complex work that demands creativity.

Top AI trends for 2023

1. Generative AI will continue to lead the way

The reason why generative AIs, like ChatGPT and digital art generator (DALL-E), are popular is because they are designed to use data to create content rather than simply analyze it. And in a world where content is king, they make it so much easier to churn out content.

While AI may remove some low-level jobs, it creates tools that businesses and creatives can use to eliminate some early grunt work, e.g., using ChatGPT to build a framework for a whitepaper, or using DALL-E to create sample ideas to help shape the direction of a creative brief.

GPT-4, the next generation of ChatGPT, was launched on March 16, 2023 to Plus users only, and its capabilities far surpass those of GPT-3.5, e.g., multi-modal (text and image processing), plugins to access the internet, and more accurate problem solving.

2. Deep learning will help shape the future of business

Diplomacy, the board game where players must actively negotiate with each other to gain the upper hand, has no random component to the game, you just need to achieve numerical superiority and another player’s support to attack.

For AI, it’s a tough nut to crack, but Meta’s AI Cicero out-competed 90% of humans in a web diplomacy tournament.

While historically, AI has excelled at strategic gameplay, this was the first example of an AI understanding open-ended negotiation and competing with humans in a difficult playing field.

With generative AIs improving, it’s also likely that AIs like Cicero will develop and be increasingly applied to gaming, business, and even negotiations.

3. AI ethics will be addressed

Was ChatGPT built on data collected without creator consent? Is ChatGPT able to comply with GDPR responsibilities?

These concerns, and others like them, have led to the concept of building AI as a glass box that explains how the AI reached its conclusion, instead of a black box. Slimmer AI tests using an explainable boosting machine (EBM) model, developed by Microsoft, found it provided comparable results to a black box model, but better explained how it got there.

While possible ethical concerns and dangers should be addressed, the potential of AI to improve the world and our lives makes it worth further exploration, albeit with caution.

4. Deep learning will help brands navigate the cookieless future

When Google retires third-party cookies in 2024, there will be a shift towards privacy-first data handling. As a result, advertisers will lose access to many of their structured datasets rendering machine learning solutions even less useful.

Deep learning solutions will be better placed to work with very large, unstructured datasets and enable advertisers to continue reaching consumers with relevant ads, while also protecting their privacy.


Deep learning is an essential tool for advertisers, but it’s impossible to know how many of these AI trends will become reality. However, it's crucial for advertisers and brands to be up to date with the new and emerging technologies and options available.

And with such a rapidly evolving marketplace, it's best to work with a trusted partner that is innovative and will help brands to thrive in the future - such as RTB House.

Deep Learning AI Digital Advertiisng


Industry insights

View all
Add your own content +