Superintelligence. When can strong AI appear?

How close are we to its creation? No one knows the exact answer, but there are various predictions from experts in the field of artificial intelligence. Opinions vary considerably, from those who believe that AI/AGI may already appear within the next few years to those who talk about a decade or more. Let’s try to understand further in the article…

“We know how to build the USA” – OpenAI

OpenAI, a company known for its ChatGPT, has been openly emphasizing its mission to create an AGI. Its leaders are optimistic about the timeline. In his address, Sam Altman noted that the team is “confident that it knows how to build AGI” and expects the first minimal signs of such intelligence to appear this year. In other words, OpenAI believes that within the next few years, we will see systems that will be able to perform a significant part of human-level work tasks. Moreover, they are already thinking about the next stage – superintelligence – and believe that its creation is possible shortly after achieving AI. They recently released a global update of GPT 5, but Altman explicitly stated that version 5 does not meet the criteria of AI because it lacks the key feature – the ability to learn in real time without restarting the system. Ultimately, OpenAI describes GPT-5 as “a significant step on the road to AGI,” but not the end of it. Then again, Sam Altman jokingly said in an interview that he expects the US to have AI “by the end of Trump’s second term” – that is, in fact, by 2028.

Google DeepMind predictions

Further, Google DeepMind, another AI company, is also engaged in AI research. Demis Hassabis, CEO of DeepMind, recently expressed the opinion that general artificial intelligence could appear within 5-10 years. Interestingly, Hassabis himself called this forecast rather conservative compared to some colleagues. According to him, some in the industry estimate that AI can be created even faster. For example, Dario Amodei, CEO of the Anthropic startup, assumes that the first human-level systems will appear in 2026.

And as a result…

Of course, these dates are nothing more than predictions and should be taken with caution. In the past, predictions about artificial intelligence have often been overly optimistic. Recent successes (e.g., neural networks that write code, engage in dialogues, and pass human-level tests) have convinced many experts that we are close to a quantum leap. Surveys of AI researchers show a trend where a significant number of experts expect AGI to appear in the middle or second half of the 21st century. For example, some surveys predict around 2040-2050. But let us repeat: the uncertainty is very high. A strong AI may not be created by the end of the century, or vice versa, it may appear in some limited form in a few years.

It is also important that there is no clear consensus on how to define the emergence of AI. Some consider the criterion to be the ability of AI to perform any economically important task better than humans, while others, such as Demis Hassabis, say that AI will really prove itself when it can make scientific discoveries and explain new phenomena on its own (for example, rediscovering the theory of relativity with only the data that Einstein had). In any case, if we see a machine that understands and solves a wide range of tasks as well as humans, this will be the birth of a strong AI.

Therefore, there are many forecasts, and they vary. The general trend is that AGI is no longer a distant fantasy and looks like an achievable goal, perhaps even in the coming decades. This means that we should already be thinking about the consequences and be prepared for its emergence.

The main point to take away is that a future with strong AI is up to us. As one of the OpenAI documents says, superintelligence could become the most powerful technology in history, with huge pros and cons. Our task is to harness this power for good, to guide its development, and not to let it run its course. Perhaps we will have opportunities to solve the long-standing problems of humanity: to overcome poverty and disease, to raise the general level of knowledge and well-being.

Натисніть, щоб оцінити цю публікацію!
[Всього: 0 Середнє: 0]

Leave a Reply

Your email address will not be published.