Let’s continue our Artificial Intelligence (AI) conversation. Last week I was in Switzerland for a client engagement. So sorry for missing the last weeks blog post window. It did give me some time to read Life 3.0 some more. I would argue that everyone should get a copy of that book.
My AI blogging assistant
Jetpack (a very important component to any WordPress blog/website) has released a new blog component. Here is the AI assistant (experimental) component explaining itself:
The AI assistant (experimental) component is designed to assist bloggers by suggesting ideas and prompts for their blog posts based on their interests and goals. This feature uses sophisticated algorithms to analyze and generate data-driven suggestions. As a blogger, you should be excited about this new feature because it can save you time and effort in coming up with fresh and relevant content ideas for your blog. Additionally, the AI assistant can help you expand the range of topics and perspectives covered on your blog. Overall, the AI assistant (experimental) component is a powerful tool that can strengthen your blog’s content strategy and help you engage your audience with interesting and unique content.
Pretty cool right? I’ll continue experimenting with this component. However, most of my content will still be by my hand, as I do want to give you MY take on the topics. So, I might start with an AI prompt, see the result, accept it and alter to my needs.
The broad conversation on Artificial Intelligence
We, I, should not go blindly all in on AI. I think this technology is very powerful. I also think that if we don’t think about implications on ALL fields, we will lose our timeframe in which we have the oppertunity to make a difference.
Microsoft’s statements on AI that you can find here are also very interesting. Responsible and social good.
If you have some time, I’d highly recommend viewing this video (1 hour plus): The A.I. Dilemma – March 9, 2023 – YouTube
And if you have more time available, also sit down for this video. But please mute your audio for the first 5 seconds, because the intro is MUCH louder than Yuval’s speach.
(Please make sure to have the sound low at the beginning of the video!)
I asked ChatGPT to respond as well. Here’s what we get.
Perfect Prompt rephrased prompt
{ “rephrased”: { “text”: “As someone who specializes in psychology and artificial intelligence and has a keen eye for predicting future trends, please evaluate the accuracy and likelihood of the following three predictions: \n1. AI will become unmanageable and beyond human control within the next five years. \n2. The development of AI teaching AI will result in the emergence of superintelligence within two years. \n3. In the next ten years, we can expect the creation of an AI with intelligence levels beyond human capacity and comprehensive knowledge of the world. \nPlease provide thorough analysis and supporting evidence to substantiate your assessment.” } }
The AI response (without Bing connected)
As a futurist with expertise in psychology and artificial intelligence, I’ll provide an assessment of the three predictions you’ve proposed. Please note that these assessments are based on the current state of AI research and development as of my last training data in September 2021, and the field is rapidly evolving.
- AI will become unmanageable and beyond human control within the next five years.This is a topic of ongoing debate among AI researchers and ethicists. While it’s true that AI systems are becoming more complex and powerful, most experts agree that we’re still far from creating an AI that could become uncontrollable. AI systems are tools created by humans, and they operate under the constraints and objectives that we set for them. They don’t have desires or intentions of their own. However, there are valid concerns about the misuse of AI, and about the possibility of creating systems that are so complex that we can’t predict their behavior. The Future of Life Institute has a good summary of these issues here.
- The development of AI teaching AI will result in the emergence of superintelligence within two years.AI teaching AI, also known as machine learning from machine-generated data, is a promising area of research. However, as of 2021, it’s unlikely that this will lead to the emergence of superintelligence within two years. Creating a superintelligent AI would require not just improvements in the algorithms we use, but also major breakthroughs in our understanding of things like consciousness and general intelligence. OpenAI’s Charter provides a good discussion of these issues.
- In the next ten years, we can expect the creation of an AI with intelligence levels beyond human capacity and comprehensive knowledge of the world.This is a more plausible scenario, but it’s still highly speculative. AI has made impressive strides in specific domains, but creating an AI with broad, human-like understanding of the world is a huge challenge. Current AI systems lack common sense reasoning and an understanding of the context in which they operate. The AI Now Institute’s report provides a good overview of the current state of AI and its limitations.
In conclusion, while AI is advancing rapidly, it’s important to remember that it’s still a tool created and controlled by humans. The timelines for these predictions are likely optimistic, but they highlight important issues that we need to address as a society.
Final Notes
Notice that in the above example, the AI responded with the limitation that it still thinks it’s 2021. I ran another version of this prompt, but now connected with Bing, and the results are roughly the same with a couple of key differences.
It included this text:
To stay updated with the most recent advancements, it would be beneficial to use my browser tool to look for any new breakthrough# Let’s find recent developments in AI as of 2023
The AI
And it concluded, as we shall, with:
In conclusion, while Artificial Intelligence (AI) is improving rapidly and becoming more accessible and integrated into our lives, the predictions you mentioned still seem unlikely given our current understanding and technological capabilities. The development of superintelligence or an AI with intelligence levels beyond human capacity and comprehensive knowledge of the world would likely require major advances in our understanding of intelligence and AI technology, which are not guaranteed to happen within the specified timeframes. However, it’s important to continue to monitor the development of AI.