Artificial intelligence was one of the defining topics on global stock markets in 2023. There are likely to be some interesting developments in this regard in the new year, say several Stanford scientists.
• Stanford scientists believe AI has not yet reached its peak
• AI is likely to bring new agents and support knowledge work
• Concerns about potential harm from deepfakes
AI – which usually refers to applications based on machine learning, in which software sifts through large amounts of data for matches and draws conclusions from them – was on everyone’s lips in 2023. Companies invested huge sums in this future technology. The possible applications are diverse: For example, such programs can evaluate images from computer tomographs faster and with greater accuracy than humans. Self-driving cars also try to predict the behavior of other road users. And chatbots or automatic playlists from streaming services also work with AI.
A lot of work will change
But AI is far from reaching its peak, say several Stanford HAI lecturers and fellows. In fact, they also foresee numerous interesting developments for 2024. One of the most important is certainly the influence that AI will have on the world of work.
According to her, mass AI adoption by companies is likely to impact knowledge workers. The work for creative people, lawyers, finance professors and others will change in 2024. Stanford scientists largely view this as positive because it would improve work and make it possible to do new things that couldn’t be done before. Only in rare cases will a job be completely automated – instead, the options can usually be expanded.
New helpful agents
If 2023 was the year you could chat with an AI, 2024 will likely see agents doing things for us like making reservations, planning a trip, or connecting to other services.
Scientists also expect steps towards multimedia, even if this will take more than just a year. While the focus so far has been on language models and image models, at some point there will probably be enough computing power available to also process videos.
Beware of deepfakes
However, a danger could arise from the expected large new multimodal models, particularly in video generation. As a result, there will certainly be more disinformation and fakes in 2024. Videos will spread of people “saying” things they never actually said. The scientists warn that consumers and voters must therefore become more vigilant against serious deepfakes.
GPU shortage
Because there is currently a run on GPU processors, scientists are worried about a global shortage of the specialized processors that run many AI applications. Especially since there are only a few companies that make them. The race between companies and states for leadership in AI is not only creating enormous pressure to expand GPU production, but also on innovators to develop hardware solutions that are cheaper and easier to manufacture and use.
Regulation in view
When it comes to AI, opinions sometimes vary widely. While some praise the enormous potential, others warn about the possible dangers of this technology. Given the importance of this topic, political decision-makers are now also seriously considering the regulation of AI.
“Digital transformation has huge potential,” said Volker Türk, the UN High Commissioner for Human Rights, in Geneva in mid-December. “But let’s face it: from artificial intelligence to neurotechnology, cybercrime, surveillance and bioweapons, we face a world in which the very foundations of humanity – human dignity and human agency – are at risk,” he warned.
The European Union (EU) has already taken action and agreed on the world’s first rules for AI in December. The proposed regulations establish obligations for AI based on its potential risks and impacts. AIs that have significant potential for damage to health, democracy, the environment or security are classified as particularly risky. Some applications will even be banned entirely, such as biometric categorization systems that use sensitive characteristics such as sexual orientation or religious beliefs.
In the US, Congress introduced the bipartisan CREATE AI Act in July 2023 to give students and researchers access to AI resources, data and tools. It remains to be seen whether Congress will act in 2024 and pass legislation like the CREATE AI Act. However, the fact that 2024 is an important election year in the USA could paralyze the legislative process somewhat.
Editorial team finanzen.net