news 
Attaining AI Superintelligence Straight On Rather Via Intermediary Stepping-Stones
In the rush to develop Artificial General Intelligence (AGI), we are often too quick to overlook the grand prize – Artificial Superintelligence (ASI). Some argue that we should focus on AGI, claiming it’s a stepping-stone towards ASI. However, I believe that this approach is misguided and may lead us down a path of missed opportunities.
The notion of going for AGI as a means to achieve ASI is akin to aiming at the wrong target. Instead of shooting straight for the moon (ASI), we are settling for circling the earth (AGI). This thinking only delays our potential achievement of the ultimate prize – ASI.
While some proponents argue that pursuing AGI will automatically lead us to ASI, this may not be the case. In fact, it’s possible that we could get stuck in a prolonged and fruitless struggle with AGI while missing out on the enormous benefits that ASI would bring.
On one hand, I acknowledge that there are risks associated with ASI. We have no idea whether we can control it or align it with human values. However, I firmly believe that by aiming straight for ASI, we may be able to avoid these issues altogether.
In contrast, our risk-averse stance and hesitation to take the leap towards ASI means that the untold benefits of this technology are either pushed into the future or entirely waylaid. By settling for AGI, we are putting a brake on progress.
Furthermore, I argue that the potential economic and societal changes brought forth by AGI would be exponentially heightened if we were instead to pursue ASI directly. For instance, will humans no longer need to work for a living? How will AGI and humanity balance the use of always-available human-level intelligence? These concerns are amplified when considering ASI.
Moreover, I believe that it’s crucial to recognize the dual-use properties of AI (conventional AI, AGI, or ASI). Any type of AI has the potential to be used for both good and ill. This is a reality we must acknowledge and confront head-on.
In conclusion, I urge us to reconsider our approach and aim straight for ASI instead of relying on intermediary stepping-stones like AGI.
Source: https://www.forbes.com/sites/lanceeliot/2025/04/01/attaining-ai-superintelligence-straight-on-rather-via-intermediary-stepping-stones/
