By Jim Lundy
European AI Law: 3 Reasons Why it Misses the Mark
Europe just passed one of the first sets of AI regulations that is getting lots of positive press. However, it is a popular trend to try to portray tech companies as evil and the reality is that AI is mainly about automation.
In this blog, we discussed three reasons why the AI law in Europe misses the mark.
To Regulate AI, You Must Understand It
At first glance, it’s not clear that the purpose of the AI law is intended to stop some of the bad cases that could be developed using AI. We will say that it does tackle the issue of Deep Fakes, but there is much more to this issue of regulating AI.
Instead, the legislation seems to focus more on the technology itself. It also implies that more computing power means AI is more dangerous, which is simply not true.
Focus on Use Cases—Not the Technology
To us, the main use case that the EU focused on was deep fixed, and there is merit to that. However, by focusing on the technology of AI, and not the use cases, these gaping holes in the regulation itself.
Deep learning models can run on a MacBook Pro and can be designed to do very bad things, like hacking. Given the focus on Gen AI and processing, this example would be able to skirt the EU Regulation as written.
If we discuss large language models (LLMs), the assumption is that you must have large processing power to run them. With the advent of small language models (SLMs), running those models on things such as cell phones and other edge devices, you don’t need processing powers to run some of those applications. This is another example where the EU law could be tightened significantly.
The AI Law Will Stifle AI Innovation in Europe
It certainly was raised as an issue, but it is becoming clear that the new AI law may cause some Asador in Europe when it comes to starting a company to develop new AI technology. Europe is currently behind in AI innovation, and this does not help its prospects.
The way to handle this would be to exempt companies that have revenue of under 10 million Euros and only have the law apply when certain thresholds are met.
The AI Law is an Attempt to Tax Large Global Companies
Many outside of Europe will just view the new AI law as a way to tax companies that do development outside of Europe. Additionally, most of the revenues of large companies that are doing AI development today do not generate a vast amount of revenue from AI. So, this is a grab by the EU and it should be reviewed and addressed.
Aragon Research proposes modifying the law to tax 7% of AI revenue versus the 7% total of revenue. It takes away some of the threat of the law itself, but it would also make it fair.
Bottom Line
The AI Law by the European Union is a good first attempt to tackle Deep Fakes, but for the rest of its intended purpose, it misses the mark. Focusing on the AI use cases beyond Deep Fakes is what is needed.
Follow the Money: Modernizing Enterprise Content in the Age of AI
Thursday, March 28th, 2024 at 10 AM PT | 1 PM ET
Follow the Money: Modernizing Enterprise Content in the Age of AI
In today’s landscape, numerous enterprises confront a pressing dilemma: their aging content management systems still harbor valuable assets that could be harnessed.
Join Aragon CEO and Founder, Jim Lundy, and Analyst, Adam Pease, in this webinar as they delve into the imperative of modernizing content repositories, transforming them from mere silos into revenue-generating hubs. Discover the key lies in Content AI.
Key topics for discussion include:
- The key trends driving the evolution of Content.
- What is Content AI and how will it help the enterprise with Revenue?
- How can enterprises gain a competitive advantage by automating and modernizing their content.