The next AI advancements may add decades to the timeline, but ultimately accelerate us towards a singularity

There have been recent breakthroughs in LLMs (Multimodal AI, boundless/cheaper training), we may witness Meta, Google, Anthropic, and Microsoft “slow down” or even reduce the pace of AI advancements in the next 5-15 years for the following reasons:<br><br>&#x200B;<br><br>1. **Global Regulations**: Threats of global regulations (e.g., EU, US) may slow down the pace of public AI advancements as companies may limit the AI capabilities they release to the public. These regulations may include restrictions on AI model sizes, the data used for training, and the domains in which they can be applied. Governments may want to have far more control over AI rather than letting it be a free for all. EU is the first one starting to regulate AI. <br><br><br>2. **Company Monopolization of AI**: Companies may choose to limit public AI capabilities and keep it for themselves, using the public as guinea pigs by slowing down public AI developments. This might accelerate AI research in secret as companies may have more freedom to experiment and advance AI beyond what’s safe to release to the public. However, this may not happen if there is enough competition in the field of AI. We may see AI arms race[s] between companies. <br><br><br>3. **Reduced Risk of Leaked AI**: Companies may “slow down” public AI advancements to reduce the risk of leaked AI models. It may become much more difficult to get leaked models, but it also slows down public AI advancement. <br><br><br>4. **Limited Computing Power**: The next boundless AI may also require boundless computing power. As we are getting close to the theoretical limit of computing power posed by physics, there may be a slowdown in AI advancement due to the lack of computing power.<br><br>&#x200B;<br><br>However, in the long run, companies and governments will not be able to slow down or stop AI development. The AI revolution is not stopping or slowing down, companies and governments may try to limit public AI advancements, but they will keep accelerating AI development in secret.<br><br>AI may be slowed down, but once we have boundless AI, boundless training data, and boundless computing power, humanity is in for an exciting future.