Today it feels like we’re living in a science fiction movie. Things that were unreal a year ago have become a reality. As the world is changing, there’s a growing conversation around technological boundaries, extending even into the world of online gaming and platforms that operate outside the UK’s self-exclusion scheme. Some believe that artificial intelligence will start the fourth industrial revolution. With this in mind, this article will take a closer look at how current AI trends may develop in the future, including potential impacts on the digital gambling industry. The trends discussed are based on perspectives from researchers, AI experts, and pioneering AI companies.
Generative AI Continues at Lightning Speed
In November 2022, generative AI took off in a way we’ve hardly seen before. It started with ChatGPT. Since then, it has only grown, and thousands of AI tools have been launched. We have compiled an overview of the 20 best AI tools here.
What Is Generative AI?
A generative AI is an AI tool that creates something new. It can create text, images, video, music, voices and so on…
Generative AI models are fed billions of images, videos or words. They learn what the difference is in the pictures, words or videos.
The model then uses its knowledge as a basis for creating something new.
There are a number of AI tools that are generative. Some examples are DALL-E for images, ElevenLabs for voices, and Stable Audio for music.
5 Types of Generative AI
Generative AI is a collective term for various tools. Here are the most popular types:
- AI text: Language models can create everything from job applications to jokes and stories.
- AI Images: Image generators can create realistic, artistic, and creative images.
- AI Voices: Now it is difficult to distinguish AI voices from real voices.
- AI Video: Video generators are getting better every week.
- AI music: Now music is made by Drake, Frank Sinatra, and so on.
But generative AI is not just entertainment.
2. AI Is Integrated into Even More Apps
By 2024, a bunch of apps will likely integrate AI. There are several reasons why this will happen:
- Developers can connect AI tools like ChatGPT to their apps (via APIs).
- Investors are looking for opportunities to exploit AI commercially.
- App developers get more capital to invest in AI.
- Users demand services with AI capabilities.
Therefore, app companies that ignore AI will risk falling behind. Here are some examples of apps that have already integrated AI:
In 2023, Snapchat launched AI-generated snaps for those who have Snapchat+.
Plus subscribers can create and send AI-generated images. Users can choose to get ideas from templates or to enter their own suggestions.
Among other things, Microsoft has integrated the image generator DALL-E 3 in Paint (for Windows 11).
At the same time, they have integrated ChatGPT into the Bing search engine (which they now call CoPilot).
Canva is an app for creating graphics and editing images. One can create infographics, social media posts, and so on.
3. Automated AI agents are coming
AI agents are programs that can perform tasks on their own. They are also called “intelligent agents”.4
They can be digital programs or physical robots.
A well-known example is the AI that you fight against in computer games. It makes its own judgments to try to overcome the player.
AutoGPT: Automatic ChatGPT
The most interesting development is AutoGPT. It is an AI agent that appeared in 2023.5
AutoGPT builds on language models such as ChatGPT by giving them autonomy.
ChatGPT only responds when it receives a question or an instruction. In contrast, AutoGPT takes initiative on its own.
How AutoGPT works:
You give the AI agent a target:
- The agent calculates what to do.
- The agent creates a list of tasks.
- Then it collects information.
- It then solves the tasks one by one.
- In some cases, the AI can outsource tasks to other AIs.
In 2024, we will likely see more companies and organizations adopting AI agents to automate work tasks.
Researchers Created a Village With AI Agents
In an interesting experiment, researchers from Stanford University tested how AI agents work in a game.
They created a virtual world called “Smallville”, inspired by The Sims. 25 game characters live here, which are controlled by ChatGPT and the researchers’ own code.
“Generative agents wake up, make breakfast and go to work. Artists paint, while writers write. They form opinions and start conversations. They reflect on their day while planning the next day,” the researchers write in the article Generative Agents: Interactive Simulacra of Human Behavior.
In the article, the researchers say that unexpected behavior occurred in the simulation. These are behaviors that were not coded in advance:
- Information dissemination: Agents share information that spreads in the city.
- Memory: The agents remembered and mentioned previous conversations.
- Coordination: They planned and attended a Valentine’s party.
Although 12 agents heard about the Valentine’s party, only five showed up. Three said they were too busy, and four agents simply did not show up.
This shows how unexpected situations can arise in the virtual world.
4. AI Influencers Become Commonplace
AI influencers are personalities who produce content and engage followers just like human influencers. In 2023, Meta invested millions of dollars in developing AI influencers. Now they’ve made AI influencers of Snoop Dogg, Mr. Beast, Kendall Jenner, and many more.
However several fans have reacted negatively to the AI celebrities.
Nevertheless, this trend does not seem to be stopping.
5. Deepfakes Become a Societal Problem
Deepfakes are fake videos, images or audio recordings that have been manipulated using AI.
They can show people saying or doing things they have never said or done.
Some examples of what deep fakes can be used for:
- To spread false political speeches or interviews
- Making fake news videos
- Creating false evidence for court cases
The problem with deepfakes is that they are difficult to detect.
To solve this problem, several universities and companies are working with technology that detects deepfakes. Then we are talking about researchers at MIT, Intel, and others.
Intel claims that its platform can detect deepfakes with 96% accuracy.