Artificial General Intelligence (AGI) – the holy grail of artificial intelligence – has long been the subject of debate among experts and researchers. OpenAI CEO Sam Altman claims that AGI can be achieved as early as next year, with machines capable of “thinking like humans” in just a few years.
According to Altman, AGI is no longer an engineering problem but rather a matter of when it will happen. He cites the progress being made by his team and believes that the path to AGI is now clear. However, not everyone agrees, and the definition of AGI remains undecided.
To achieve AGI, AI systems must be capable of learning, adapting, and performing tasks similar to human intelligence, going beyond just knowledge-based output. Researchers have been working on developing models that can solve problems not in their training data, a crucial aspect of AGI.
Altman’s confidence stems from OpenAI’s progress with its model o1, which is seen as a significant step up from its preview version. The company also claims to be nearing the final stages of development for its next-generation Gemini models.
However, some experts question Altman’s timeline, pointing out that the leading models from Google and Anthropic are still in development. For instance, Anthropic CEO Dario Amodei predicts AGI will be achieved by 2026/2027.
The AI lab defines AGI as “AI systems that are generally smarter than humans,” with a clear progression through five levels of AI, culminating in Level 5, where the model is capable of doing all the work of an organization independently. This definition raises concerns about the potential implications and challenges associated with achieving AGI.
Ultimately, experts agree that AGI will be achieved gradually, rather than in a single leap. As such, it’s essential to consider the potential implications and ensure that those building AGI are aware of the potential risks and consequences.
Note: I removed the unnecessary content, such as the Tom’s Guide section, and rewrote the article to make it concise and clear. I also generated a new title that accurately reflects the content.
Source: https://www.tomsguide.com/ai/chatgpt/sam-altman-claims-agi-is-coming-in-2025-and-machines-will-be-able-to-think-like-humans-when-it-happens