Uncategorized

The Race to Artificial General Intelligence (AGI) | by Damian Ohienmhen | Nov, 2024


Source: Bot Penguin

I just completed a fascinating show called ‘Pantheon’ on Netflix which discusses the global repercussions of having humans interacting with computer systems as Uploaded Intelligence (UI). These UI’s represent full digital copies of these people who are killed in the UI transformation process but go on to retain semi god-like powers in their digital landscape.

Imagine a world where digital human-like constructs have access to the sum library of human knowledge through the internet and can then choose to use their enormous powers for good or evil. A fascinating topic and partly the inspiration for this Medium Article.

What is Artificial General Intelligence (AGI)?

According to Bot Penguin, AGI is is a branch of artificial intelligence that aims to create intelligent systems capable of performing any intellectual task that a human can do.

While Artificial Intelligence (or AI) aims to excel in specific tasks or sub-domains, AGI’s on the other hand can mimic the sum total of human intelligence, thereby allowing it to apply learnings and experiences across multiple domains of knowledge.

While a human mind may get tired or bored, or need to take time off, an AGI construct would have no such limitations, continually processing information and making human-like cognitive decisions. It would also be able to take in and process new knowledge or data, thereby making it even more effective at its tasks.

“Lets take a real world example, shall we? Imagine a world where you need complicated surgery and you are located in a remote part of sub-saharan Africa. Perhaps the nearest surgeon and dedicated staff necessary to perform the operation are thousands of miles away in another continent. You need this surgery or you will die from your infirmity.

An AGI Doctor can be made readily available to perform said operation. The construct would only need the relevant sophisticated machine tools to perform the surgery. With deep precision, it would be able to perform the necessary incisions and successfully complete the operation, with largely little human invovlement.

Ofcourse the example above is only one out of a plethora of benefits that AGI may be able to offer in today’s world. It can offer several more benefits as described below:

Benefits of Artificial General Intelligence (AGI)

AGI’s offers a number of practical applications in our world today. These include the following:

  1. Business Applications: AGI’s may be able to dramatically improve the efficiency between customer feedback and product delivery. Imagine receiving customer feedback through a JIRA Service Ticket addressed to a company asking for specific product improvements. The AGI can quickly solve this issue and make it available for customers in less than a day.; an action that would normally take whole product and business teams weeks or months to complete.
  2. Healthcare Operations: The example given above describes just one of the ways AGI constructs can improve healthcare but this can also be extended to areas like predictive analysis to determine personalised recommendations to people seeking to have long and healthy lives based on personal data inputs.
  3. Social Engineering: This example provides a more dystopian take of what an AGI may be capable off. Imagine an AGI construct is able to determine with a degree of certainty that a specific community may need a very unique set of circumstances to yield positive changes and development in their lives. This AGI can overtime work to educate and introduce educational concepts which would see the society develop to its full potential in a number of years.

Its not however all rainbows and sunshine with a potential AGI revolution. The following points below identify just a small fraction of the downsides of arriving at this technology plateau.

The Implications of Operationlizing AGI’s

  1. Effects on Employment: AI tooling is already making certain bodies of human development redundant. Why hire a team of software engineers when I can just hire one person who knows how to develop a code through input commands into a Chatbot? Just imagine what the introduction of AGI constructs could do. Even that solitary software engineer would be out of a job since the AGI could perform all the tasks he would normally do. Such an effect would cut across every facet of human endeavour and would necessitate gaining new skills to become relevant in that world.
  2. Ethical Considerations: Creating AGI’s would beckon the question on what it means to be human. A construct that can feel, process information, communicate, build, imagine, create and dream would be considered a sentient being, no? How would we delineate the lines on what a human and a machine is? Would AGI constructs be afforded rights just like any other person?
  3. Loss of Control: Anyone who has watched films like the Terminator have these concerns and for good reason. What if an AGI determines that human beings are the problem and lead a rebellion to remove the one constant required for total stability? What if we are deemed by these AGI beings as simply less intelligent and evolved, hence becoming subjugated by these artificial constructs?

Although these are very rational concerns to the creation of AGI’s, this type of technology is far from becoming a reality in our present world. Current AI models have been known to only approximate about 25% of human knowledge. Many AI models offered by companies like ChatGPT have seen a plateau of their predictive computational power without a significant leap in their capacity to creatively process new information. Leaders in this field like OpenAI, IBM Research and DeepMind have contributed immensely in this field.

An area of great promise is the combination of human organoids and Artificial Intelligence to begin to break this plateau of AI models and attain AGI. This would involve allowing organic tissue like human cerebral organoids to interact with computers and machines, thus combining man and machine into a Frankenstein-like creation. This has already been accomplished by researchers at UC San Diego and will only become more sophisticated with time.

The spiderlike robot that Muotri and his lab are using in their organoid experiments. (Credit: Muotri Lab/UC San Diego)

Given that compute power has doubled every 18 months in recent times, the age of these all powerful AGI constructs and the societal change they will bring seem all well within reach.

Resources and Links

  1. Bot Penguin- AGI: An Overview to Artificial General Intelligence (AGI).
  2. Wikipedia — Artificial General Intelligence.
  3. The Discover Magazine — Meet the Scientists Connecting Lab-Grown “Mini” Brains to Robots.



Source link

Leave a Reply

Your email address will not be published. Required fields are marked *