Is an AGI even required to achieve similar results?

A Comprehensive Artificial Intelligence Services technical report model by Drexler, from 2019, seems useful to revisit at this time. Instead of focusing on the hypothetical scenario of a single superintelligent agent that surpasses human intelligence, we should, the report argues, consider the more realistic possibility of a diverse and interconnected network of AI systems that provide various services for different tasks and domains. They call this approach Comprehensive AI Services (CAIS).

The main advantages of CAIS are that it avoids some of the conceptual and technical difficulties of defining and measuring intelligence, and that it allows for a more fine-grained and flexible analysis of the potential benefits and risks of AI.

It's also a good way of considering where we have arrived at, with AgentGPT's operating via Hugging Face or via AutoGPT for example. By connecting a range of Narrow AI tools to perform the tasks that they are optimised for, and having a 'manager' assign the allocation of these tasks, giving the correct prompts for each agent, this 'comprehensive' approach could provide similar results to an AGI?

The authors of the technical report suggested that CAIS can help us better align AI systems with human values and goals, by enabling more human oversight and collaboration, and by fostering a culture of responsibility and accountability among AI developers and users. Which seems far more plausible than trying to do that with a monolithic AGI.

The authors conclude by outlining some of the open questions and challenges that CAIS poses for AI research and governance, such as how to ensure the reliability, security, and interoperability of AI services, how to balance the trade-offs between centralization and decentralization of AI systems, and how to promote ethical and social norms for AI use and development. These are questions that exist for all AI systems.

Comments

The AI Dilemma and "Gollem-Class" AIs

From the Center for Humane Technology Tristan Harris and Aza Raskin discuss how existing A.I. capabilities already pose catastrophic risks to a functional society, how A.I. companies are caught in a race to deploy as quickly as possible without adequate safety measures, and what it would mean to upgrade our institutions to a post-A.I. world. This presentation is from a private gathering in San Francisco on March 9th with leading technologists and decision-makers with the ability to influence the future of large-language model A.I.s. This presentation was given before the launch of GPT-4. One of the more astute critics of the tech industry, Tristan Harris, who has recently given stark evidence to Congress. It is worth watching both of these videos, as the Congress address gives a context of PR industry and it's regular abuses. "If we understand the mechanisms and motives of the group mind, it is now possible to control and regiment the masses according to our will without their

Beware the Orca, the challenge to ChatGPT and Palm2 is here

So Google's 'we have no moat' paper was correct. If you train an LLM wisely then it's cost effective and cheap to produce a small LLM that is able to compete or even beat established, costly LLMs, as Microsoft has just found. It's another excellent video from AI Explained, who goes through some of the training procedures, which I won't get into here. Orca, is a model that learns from large foundation models (LFMs) like GPT-4 and ChatGPT by imitating their reasoning process. Orca uses rich signals such as explanations and complex instructions to improve its performance on various tasks. Orca outperforms other instruction-tuned models and achieves similar results to ChatGPT on zero-shot reasoning benchmarks and professional and academic exams. The paper suggests that learning from explanations is a promising way to enhance model skills. Smaller models are often overestimated in their abilities compared to LFMs, and need more rigorous evaluation methods. Explana

What is happening inside of the black box?

Neel Nanda is involved in Mechanistic Interpretability research at DeepMind, formerly of AnthropicAI, what's fascinating about the research conducted by Nanda is he gets to peer into the Black Box to figure out how different types of AI models work. Anyone concerned with AI should understand how important this is. In this video Nanda discusses some of his findings, including 'induction heads', which turn out to have some vital properties. Induction heads are a type of attention head that allows a language model to learn long-range dependencies in text. They do this by using a simple algorithm to complete token sequences like [A][B] ... [A] -> [B]. For example, if a model is given the sequence "The cat sat on the mat," it can use induction heads to predict that the word "mat" will be followed by the word "the". Induction heads were first discovered in 2022 by a team of researchers at OpenAI. They found that induction heads were present in

Charting the emergence of AGI?

Search This Blog