What to expect from AI in 2024

The bewildering speed with which generative artificial intelligence developed in just one year dominated the news in 2023. This set off both alarm and excitement: how AI innovation offers solutions to everyday chores, while taking us into dangerous new terrain. Geeks and tech freaks like me watched fascinated as AI applications accelerated in the past year. It was only in March 2023 that I got hooked into AI apps and wrote the front page report in Nepali Times titled Ayo AI.

Last week, this paper carried an AI-assisted artwork by digital artist Ashim Shakya (pictured below), showing the enormous possibilities in enhancing existing creative processes.

When I began to grasp AI knowhow, I never expected to continue using generative AI apps but was hooked into it. From Open AI’s ChatGPT 3.5 and two other image generators, I have now tried out over 1,000 apps. Even that is just a drop in the ocean of AI apps available today.  

In May, when Ayo AI was featured in this paper, there were about 640 AI apps. There were 10,842 relevant AIs to 9,000 tasks as we went to press on Thursday. It could impact 5,000 job sectors if these AIs are directly embedded into their workflows. 

So far, there is no evidence of this happening yet. According to the world’s number one AI aggregator, ‘There is an AI for that’ among job holders, 98% of those impacted will be the communications managers and others in this sector. But that is just an opinion, not yet a fact.

For content creators, 2023 has been a great year with generative AI app for almost every task. These can be done with just a few text prompts to generate ideas, stories, voiceovers, images, illustrations, websites, infographic designs, animations, music and videos. 

ChatGPT is no longer the dominant chatbot, but is still one of the best. It was my favourite until I started using Anthropic’s Claude-2 that was released in July. It gives 200,000 tokens, which means that we can upload large documents up to 75,000 words and have it summarised in bullet points simplifying the text. I could use it to synthesise this review, but am doing it the old fashioned way.  

The biggest sensation is Gemini AI, considered the most capable AI model developed by Google DeepMind, able to communicate and generate human-like texts. Gemini Pro Vision can interpret any image through its image reasoning task, and has the ability to describe the image and write captions. Its most significant feature is to identify if a photograph or artwork is original or AI-generated.

AI-powered image generators have advanced a lot and taken such a huge leap in creating realism in photos and art. Videos have advanced to another level.

InVideo, for example, can produce a whole video with just a text prompt: from writing a complete script, a voiceover, and creating visuals selected from its 9 million video stock. A major milestone in generative AI is the Runway, which generates amazing visuals with just a text prompt. 

Another new addition that could be Runway’s biggest competitor is Pika, its web-based version is now available but has a long waitlist. (I finally managed to get access after a two week wait.) It blew me away with its features: from text to video, video to video. And users can upscale the video, add duration, have out-painting, in-painting and a canvas feature.

AI-generated music is also gaining momentum as part of the audio-generation tools and platforms and being hugely invested by some of the biggest tech companies. Suno AI is considered the best music app so far, and was started by a group of musicians in partnership with AI experts in Cambridge, Massachusetts. 

In December, it partnered with Microsoft, which has integrated Suno AI in its Copilot as a plugin. Facebook and Google are also partnering with other music companies to create their own AI music audio generation apps.

AI is not hype anymore, it is a reality. This is the right time to be in the business. For content creators with technical knowledge and skills for design software like the Adobe products or Affinity, it enriches their work and makes them more productive, cost-effective and enhances inventiveness. Warning: it takes time to master the technique and perfect content.

 But AI-generated content is still not as lucrative as people have been led to believe. Customers still prefer original writing and art, photography or videos. YouTube, the biggest platform for creators, encourages original content.  

YouTube users are often critical of AI-generated content as they do not want to spend time watching fake content created with AI. I tried that once with my own YouTube channel and many wrote angry messages and they felt offended. 

So far, the only AI content that really sells well on YouTube is news about latest AI updates, videos with clickbait titles about how to generate income using AI, or those showcasing new AI apps, interviews with AI experts, and especially scaremongering news about AI destroying humanity.

Instagram has been a popular platform for original and creative content creators and is now crowded with millions of them. However, their AI-generated content do not get as good a response as original posts. 

There is fear mongering about how AI could destroy the careers of people in the creative fields, but AI generative tools have not ruined professional content creators like the designers, content writers, photographers, videographers and paint artists. 

It is true that the evolution of the AI generative tools has been speeding up, and we can predict even more advanced generative AI apps. The age of content creation is here and AI apps may be commercially viable and socially acceptable if used ethically and with transparency. 

For hardworking creative professionals, technology transformation has never been a threat or a challenge to their artistic creations. This argument, or rather assumption, about generative AI being a threat is exaggerated. A lot of creative people have joined the AI revolution, enhancing their original pieces of artistic creations.