During Korean Blockchain Week and Token2049 in Singapore, crypto project leaders shared their plans for utilizing AI in their projects. Rune Christensen, founder of Maker, explained that they will be relaunching their project as a collection of sub-DAOs that will employ AI governance. The AI will not enforce rules but will be used for coordination and communication. Illia Polosukhin, founder of Near, discussed the potential use of blockchain to prove the authenticity of content and differentiate between genuine content and AI-generated content. The Sandbox’s Sebastien Borget mentioned that they have been using AI for content moderation and are exploring its use for music and avatar generation. Vance Spencer from Framework Ventures outlined various use cases for AI in the crypto space, including training AI models and selling them as tokens. These insights from industry leaders provide a deeper understanding of how AI is being implemented in the crypto industry.
The article also touches on the relationship between AI and crypto. Patrick Scott from Dynamo Dao highlighted PayPal founder Peter Thiel’s view that AI is a technology of control, while crypto is one of liberation. Thiel argues that AI could potentially enable centralized control of the economy, whereas crypto offers a decentralized and individualized world. This perspective adds an interesting layer to the discussion of AI and crypto.
In another development, Roblox has introduced a new feature called Assistant, which allows users to build virtual assets and write code using generative AI. This feature aims to make asset creation and coding more accessible for users. Additionally, a study by Boston Consulting Group found that below-average workers benefited the most from using AI tools, with their output improving by 43%. On the other hand, above-average workers experienced a smaller improvement of 17%. Interestingly, workers who used AI beyond its current capabilities performed worse, as the AI provided them with plausible but incorrect responses.
Google is also preparing to release its competitor to OpenAI’s GPT-4, called Gemini. Google’s goal is to surpass GPT-4 by offering not only text generation capabilities but also image generation. This would enable the creation of contextual images. Sergey Brin, co-founder of Google, is reportedly playing a key role in evaluating and training the models for Gemini. The article also mentions that a version of Gemini has been tested and found to be on par with ChatGPT-4, with an updated knowledge base that prevents hallucinations.
In China, researchers at Tsinghua University have developed an AI system that plans cities based on the concept of “15-minute cities,” which prioritize walkability and green spaces. The AI was found to be better at tedious computation and repetitive tasks, completing in seconds what human planners took 50 to 100 minutes to do. While the AI designs showed improvements in access to services and green spaces, the increase was relatively modest. However, in a blind judging process, urban planners preferred some AI designs over others, suggesting that AI can assist humans in certain aspects of city planning.
Lastly, British actor Stephen Fry revealed that his voice was cloned by AI for a historical documentary. Fry played a clip of the narration, emphasizing that he did not record it himself and that an AI model was trained on his reading of the Harry Potter books to create the narration. This incident raises concerns about AI’s potential to replicate and repurpose the work of actors and writers without proper attribution or compensation.
Overall, these developments highlight the diverse ways in which AI is being incorporated into the crypto industry and beyond. From AI governance to content moderation and city planning, AI continues to play a significant role in shaping various sectors. However, the ethical implications and potential risks of AI cloning and misuse also warrant careful consideration.
Source link
