Author: Vince Quill, CoinTelegraph; Compiled by Tong Deng, Jinse Finance
Ilya Sutskever, co-founder of OpenAI, recently gave a speech at the Neural Information Processing Systems (NeurIPS) 2024 conference in Vancouver, Canada.
Sutskever stated that the increase in computational power through better hardware, software, and machine learning algorithms has outpaced the total amount of data available for training artificial intelligence models. The AI researcher likened data to fossil fuels, which will eventually be depleted. Sutskever said:
"Data is not growing because we only have one internet. You could even say that data is the fossil fuel of AI. It's created in a certain way, and now we're using it, and we've reached the peak of data, and there won't be any more data." "—We have to deal with the data we have."
The co-founder of OpenAI predicted that agent AI, synthetic data, and reasoning time computation are the next evolutionary directions of AI, which will ultimately give rise to superintelligent AI.
Chart comparing the computational power and dataset size of AI pre-training. Source: TheAIGRID, Ilya Sutskever
Agent AI is sweeping the crypto world
Agent AI will go beyond the current chatbot models, making decisions without human input, and with the rise of AI meme coins and large language models (LLMs) (such as Truth Terminal), agent AI has become a popular narrative in the crypto space.
After LLMs began promoting a meme coin called Goatseus Maximus (GOAT), Truth Terminal quickly gained popularity, with the coin's market cap ultimately reaching $1 billion, attracting retail investors and venture capitalists.
GOAT token market information. Source: CoinMarketCap
Google's DeepMind AI lab has launched Gemini 2.0 - an AI model that will power AI agents.
According to Google, agents built using the Gemini 2.0 framework will be able to assist with complex tasks, such as coordinating between websites and logical reasoning.
The progress of AI agents capable of independent action and reasoning will lay the foundation for AI to break free from data illusion.
The occurrence of AI illusion is due to incorrect data sets and the increasing dependence of AI pre-training on using old LLMs to train new LLMs, which will degrade performance over time.
