Looking into 2023: 3 challenges facing the AI industry
According to IDC, it is estimated that global spending on artificial intelligence will reach USD118 billion this year and exceed USD 300 billion in 2026.
While this shines an optimistic light on the industry, in the year ahead, we’ll also see some challenges centred on the global recession. Some businesses will look to double down on AI during this time, however, others will need to pull resources and slow their innovation.
Looking into 2023 and beyond, we take a look at some of the key challenges facing the AI industry, how they’ll impact businesses, and what they can do to combat them.
Training and running AI models can be an expensive matter, especially when you consider the amount of energy consumed. When you look at the amount of computing power that was used to train GTP3, it’s said that its training was equal to driving a car to the moon and back. With today’s energy crisis as a backdrop, this is not something that is sustainable.
It is therefore clear that in the year ahead the cost of energy is going to impact AI development. Throwing more data at things isn’t the right approach. Instead, companies need to look at better engineered data and how we can train models more efficiently.
This means engineers will have to be smarter with what they do. The days of running AI networks in parallel on unlimited resources have gone even for the big companies. Instead, it is about making the most of what you have now and in the future. Smart engineers and smart solutions on existing hardware is the future.
During a recession, companies often go back to their roots. So, those that use AI to fix fundamental basic problems in companies will be the ones that succeed. From a synthetic data point of view, this means companies will be looking to fix basic problems with AI — such as object detection or recognition — rather than problems requiring deep learning.
In 2023, we are likely to see some companies optimise AI, while others will halt innovation. As the recession ends, we will see who the winners and losers are. Those who have stuck to their convictions will become leaders with the rest struggling to catch up in the new AI-driven world.
A recession also makes or breaks R&D departments. These departments will need to show tangible results to the business during this period. However, more likely than not they will find themselves not knowing how to productise and deliver industrial strength AI solutions, having hopelessly underestimated the cost and effort required, and therefore fail and close down.
3. Public perception
There’s an increasing public suspicion around AI, a fear of the “big brother element”. But at the same time, people will continue to enjoy the benefits such as reduced false alarms on their video doorbells.
An increasing number of businesses and consumers in 2023 will expect “explainable AI”. This is a trend featured in a recent survey conducted by IBM, which showed that 84% of respondents said that “being able to explain how their AI arrives at different decisions is important to their business.”
It is important to note here that responsible AI will be defined by forthcoming legislation and implementing it will be costly. Because of this, companies will face a difficult challenge between the wide adoption of AI and the adoption by only those divisions who can afford it.
There is no doubt AI will still be seen as a critical component to business success in 2023. According to Deloitte’s “State of AI in the Enterprise” report, 94% of business leaders said that AI was critical to success over the next five years. The success, however, will depend on how businesses will take on and solve the challenges described, how they will make up for skill shortage, rising energy costs, coping with recession and improving public perception.
If you’re interested in finding out more about Mindtech and the role of synthetic data, please get in touch: https://www.mindtech.global/
Looking into 2023: 3 challenges facing the AI industry was originally published in MindtechGlobal on Medium, where people are continuing the conversation by highlighting and responding to this story.