Our take on 'Super Intelligence' vs 'Artificial Intelligence' terminology debate - input to Newsweek

There has been an interesting debate over the past week about replacing Artificial Intelligence (AI) with Super Intelligence (SI).

The current push followed President Donald Trump’s decision to direct US federal agencies to use “Super Intelligence” and “SI” in place of “Artificial Intelligence” and “AI”.

A few days later, Elon Musk weighed in on X: “No more AI. SI. It’s better.” He subsequently said that SpaceXAI would be renamed SpaceXSI (see Reuters story on this: https://www.reuters.com/business/media-telecom/musk-says-he-will-rename-spacexai-spacexsi-2026-10-04/)

Newsweek (https://www.newsweek.com/) approached our Centre for our perspective on whether this terminology is likely to gain traction beyond government and what Musk’s intervention might mean for the wider AI sector.

The article is now published: https://www.newsweek.com/elon-musk-grok-developer-rebrand-after-trump-ai-remark-12521992 The article quotes our Centre head:

Professor Dhaval Thakker, head of the Centre for Responsible AI at the University of Hull in the U.K., told Newsweek that Musk's endorsement gives the term its most significant private-sector advocate yet and increases its visibility beyond government circles. However, he agreed that "super intelligence" already has a specific meaning in AI research. "Calling today's AI systems 'super intelligence' risks blurring the distinction between the systems we currently have and a much more advanced capability that remains an important subject of research and debate," Thakker said. He said Musk has significant influence over public technology narratives and can rapidly move terminology from political debate into mainstream discussion. Even so, he expects AI to remain the dominant term internationally because it is deeply embedded across research, education, regulation and industry.

Here are the answers to all the questions Newsweek was intersted in:

• Does Musk's decision increase the likelihood that "Super Intelligence" gains traction beyond government circles?

  Yes, certainly. Until Musk's intervention, "Super Intelligence" could largely have been seen as terminology imposed by the US administration. Musk’s comments about adopting it gives the term a significant private-sector champion and inevitably increases its visibility.
  

  However, visibility is not the same as acceptance. "Artificial intelligence" is a deeply embedded term across research, industry, education, regulation and everyday language. More importantly, "superintelligence" already has a particular meaning within AI: it generally refers to intelligence that substantially exceeds human capabilities, rather than being a synonym for all AI.

  
  

  That creates a genuine problem. Calling today's AI systems "Super Intelligence" risks blurring the distinction between the systems we currently have and a much more advanced capability that remains an important subject of research and debate.
  

• How influential is Musk when it comes to shaping language and branding trends in the AI sector?
  

  Musk in my opinion has enormous influence over public technology narratives. Few technology leaders have the ability to introduce a phrase through a social media post and immediately have it discussed internationally

  
  

  But his influence should not be confused with the ability to determine technical terminology. Academic communities, standards organisations, regulators and other technology companies all contribute to how terminology becomes established.

  What Musk can do very effectively is accelerate a term from political language into popular and commercial discussion. Whether it then becomes the accepted terminology is another question.

  
  

• Could the term "artificial intelligence" eventually disappear, much as older technology terms have faded from use?

  
  

  Eventually, perhaps, but I do not think this particular change will make “Artificial intelligence" disappear quickly.
  

  AI is unusual because the term has existed for around 70 years and has survived several technological generations and even periods in which enthusiasm for the field declined. It is now embedded in legislation, university degrees, professional roles, companies, products and international standards.

  Terminology does change when the technology itself changes. We may therefore increasingly use more specific terms such as agents, foundation models, autonomous systems or perhaps AGI and superintelligence. But replacing "AI" wholesale with "SI" is much harder because these terms currently describe different concepts.
  

• Looking five years ahead, do you think people will still be calling this technology AI?

  
  

  Yes. I would be very surprised if "AI" were not still the dominant general term in 2031.
  

  We may see "SI" become more common in US government communications and perhaps among companies wanting to associate themselves with frontier capabilities. But internationally, and particularly in science and engineering, I expect AI to remain the umbrella term.

  There is also a practical reason for that. If every machine learning system, chatbot and recommendation algorithm suddenly becomes "superintelligence", the word "super" stops communicating anything meaningful about capability.

  
  

• Does Musk's apparent embrace of the terminology put pressure on rivals such as OpenAI, Anthropic, Google and Microsoft to respond?

  
  

  It puts some communications pressure on them, particularly because the US federal government is a major policymaker, customer and funder of technology. 

  But I do not think they necessarily need to rebrand themselves or their technologies. In fact, there may be advantages in not doing so.

  
  

  For companies concerned about responsible AI, safety and public trust, describing current systems as "superintelligence" could be seen as unnecessarily amplifying the technology's capabilities. It risks contributing to hype at precisely the point when we need clearer communication about what these systems can and cannot actually do.

  So Musk's move will undoubtedly make other technology companies discuss the terminology. Whether they follow him is a much bigger question. My expectation is that we will initially see two vocabularies operating in parallel: "SI" within parts of the US political and policy environment, while "AI" continues to dominate internationally, commercially and scientifically.

  
  

  For someone working on Responsible AI research, the bigger issue for me is therefore not whether two letters change from AI to SI. It is whether changing the language changes public expectations. Responsible AI requires us to communicate capabilities, limitations and risks accurately. Calling systems "superintelligent" before we can demonstrate that they actually are may make that task harder rather than easier.