OpenAI co-founder Ilya Sutskever emphasizes the need for innovative approaches to scaling machine intelligence to address its current limitations.
OpenAI co-founder Ilya Sutskever recently delivered a lecture at the Neural Information Processing Systems (NeurIPS) 2024 conference in Vancouver, Canada.
In his speech, he posited that the era of artificial intelligence pre-training is coming to a close and predicted the emergence of superintelligence.
The total amount of data available for AI model training is outpaced by the increase in computing capacity through improved hardware, software, and machine-learning algorithms, according to Sutskever.
The AI researcher compared data to fossil fuels, which will inevitably deplete.
OpenAI co-founder Ilya Sutskever stated:
“Data is not growing because we have but one internet. You could even say that data is the fossil fuel of AI. It was created somehow, and now we use it, and we’ve achieved peak data, and there will be no more — we have to deal with the data that we have.”
OpenAI co-founder predicted that the next evolutions of artificial intelligence would involve inference-time computation, synthetic data, and agentic AI, ultimately resulting in the development of an AI superintelligence.
Crypto World Is Ablaze With AI Agents
The rise of AI meme coins and large-language models (LLMs) such as Truth Terminal has made AI agents a popular narrative in the crypto space.
They are capable of making decisions without human input, surpassing the capabilities of today’s chatbot models.
Upon the LLM’s promotion of a memecoin named Goatseus Maximus (GOAT), Truth Terminal went viral, ultimately reaching a market capitalization of $1 billion.
This attracted the attention of retail investors and venture financiers.
Google’s DeepMind artificial intelligence laboratory unveiled Gemini 2.0, an artificial intelligence model that will fuel AI agents.
Google has stated that agents constructed with the Gemini 2.0 framework will be capable of facilitating intricate tasks, including logical reasoning and website coordination.
The foundation for AI to transcend data hallucinations will be established by advancements in AI agents that can independently act and reason.
AI hallucinations are the result of incorrect data sets and the increasing reliance on elder LLMs to train newer LLMs in AI pre-training, which results in a gradual degradation of performance.
