The development of artificial intelligence is accelerating: some scientists predict that within the next few years, machines will become smarter than even the smartest humans. Artificial general intelligence (AGI) is a field of artificial intelligence (AI) research in which scientists strive to create a computer system that is generally smarter than humans. These hypothetical systems could possess a degree of self-awareness and self-control, including the ability to edit their own code, and be able to learn to solve problems like humans, without specialized training.
The term was first used in the collection “Artificial General Intelligence” (Springer, 2007), edited by computer scientist Ben Goertzel and artificial intelligence researcher Cassio Pennachina. However, the concept has been around for decades throughout the history of artificial intelligence and appears in numerous popular science books and films, according to livescience.com.
AI services in use today, including the core machine learning algorithms used at Facebook and even large language models (LLMs) like ChatGPT, are considered “narrow.” This means they can perform at least one task—such as image recognition—better than humans, but are limited to that specific task type or set of actions based on the data they were trained on. On the other hand, AI-based AI will go beyond the data they were trained on and demonstrate human-level capabilities across a variety of domains of life and knowledge, possessing the same level of reasoning and contextualization as humans.
allsee.team
But because AI has never been created, there is no consensus among scientists about what it might mean for humanity, what the most likely risks are, and what the potential societal consequences might be. Some previously assumed it would never happen, but many scientists and technologists agree that AI is possible within the next few years, including computer scientist Ray Kurzweil and Silicon Valley executives such as Mark Zuckerberg, Sam Altman, and Elon Musk.
Artificial intelligence (AI) has already demonstrated a number of advantages in various fields, from assisting scientific research to saving time. Newer systems, such as content generation tools, create, for example, illustrations for marketing campaigns or email drafts based on user conversational patterns. However, these tools can only perform the specific tasks they’ve been trained to perform, based on the data provided by their developers. On the other hand, AI has the potential to unlock a whole new range of benefits for humanity, particularly in areas that require problem solving.
Hypothetically, AGI could contribute to increased resource abundance, accelerate the global economy, and unlock new scientific knowledge that expands the boundaries of what’s possible, OpenAI CEO Sam Altman wrote in a blog post published in February 2023—three months after ChatGPT went online. “AI has the potential to empower everyone in incredible new ways; we can imagine a world in which each of us has access to assistance with virtually any cognitive task, dramatically enhancing human ingenuity and creativity,” Altman added.
However, there are numerous existential risks posed by AGI, ranging from “misalignment,” in which the system’s underlying goals may not align with the goals of the humans operating it, to a “non-zero chance” that a future system will destroy all of humanity, Musk stated in 2023. A review published in August 2021 in the Journal of Experimental and Theoretical Artificial Intelligence outlined several potential risks of a future AGI system, despite the “enormous benefits for humanity” it could potentially bring.
“The review identified a number of risks associated with AGI, including AGI becoming uncontrollable by owners/managers, setting or developing unsafe goals, developing unsafe AGI, AGI with low ethical, moral, and value orientations, inadequate governance of AGI, and existential risks,” the study’s authors write.
The authors also hypothesized that future technology could “possess the ability to recursively self-improve, creating more intelligent versions of itself and modifying its preprogrammed goals.” The researchers also write that there is a possibility of AI being created by groups of people for malicious purposes, as well as “catastrophic unintended consequences” caused by AI’s benevolent intentions.
There are conflicting opinions about whether humans are truly capable of creating a system powerful enough to become an AGI, and about when such a system might be created. An analysis of several large surveys of AI scientists shows that the general consensus is that this could happen before the end of the century, but views have shifted over time. In the 2010s, the generally accepted view was that AGI would appear in about 50 years. However, recently, this estimate has sharply declined to 5-20 years.
In recent months, a number of experts have suggested that the AGI system will emerge as early as this decade. This is precisely the timeline Kurzweil proposed in his book, The Singularity Approaches (2024, Penguin): the moment we reach AGI will correspond to the technological singularity.
This point will be a point of no return, after which technological growth will become uncontrollable and irreversible. Kurzweil predicts that reaching this important milestone in the development of artificial intelligence (AI) will lead to the emergence of superintelligence by the 2030s, and then, in 2045, people will be able to directly connect their brains to AI, expanding the capabilities of human intellect and consciousness.
Other members of the scientific community suggest that the emergence of AGI could happen in the near future. For example, Goertzel suggested that we could reach the singularity by 2027, while DeepMind co-founder Shane Legg stated that he expects AGI to emerge by 2028. Musk also predicted that AI will surpass the smartest human by the end of 2025.
For over a decade, scientists have been creating artificial brain tissue, known as brain organoids, in the laboratory and integrating them with computer chips. New research in China has taken this idea a step further, even implanting an organoid into a humanoid robot, offering a striking glimpse into the future of this technology.
While creating brain-equipped robots is still a distant concept, researchers believe organoids could help people with neurological diseases today, reports popularmechanics.com.
A major battle is raging in the world of technology: who will win the proverbial battle: humans or artificial intelligence? But many researchers aren’t taking this “us versus them” approach, instead considering the future in a much more inclusive and, essentially, memoir-like manner, asking: why not both?
For years, scientists have been developing ways to create biocomputers using brain-like tissue (brain organoids) grown in the lab and connected to computer chips. The ultimate goal is to create a kind of hybrid intelligence, a potentially conscious entity capable of harnessing the strengths of both the human brain and artificial intelligence. If this all sounds a bit sci-fi, that’s because researchers have only recently succeeded in connecting organoids to computer chips in any meaningful way.
In 2013, scientists grew the very first mini-brain in a test tube, and since then, further research has integrated these lab-grown brains with electronics. In late 2023, researchers from Indiana University Bloomington connected their “Brainoware” architecture to an artificial intelligence tool, and now researchers from Tianjin University in China report that they have also created a robot with “organoid intelligence,” or BI. The robot, dubbed MetaBOC, is capable of obstacle avoidance, tracking, and grasping objects, and has expanded the brain-on-a-chip architecture from two to three dimensions. The results of the study were published in the journal Brain.
See also... New method for observing individual atoms confirms century-old quantum mechanical theory |
“Brain-computer interface on a chip is a technology that uses in vitro-grown “brains” (e.g., brain organoids) in combination with an electrode chip to achieve information interaction with the outside world through encoding and decoding, as well as stimulation and feedback,” said Min Dong of Tianjin University in a press release translated from Chinese.
The result is a robot that’s half brain, half electronics, and 100% cursed. The grapefruit-sized, putty-like organoid resides in the head of a bipedal humanoid robot, offering a dysfunctional yet strikingly realistic vision of where this technology could lead—but the road to that vision is fraught with obstacles. In an interview with the South China Morning Post, one of the Tianjin researchers noted that “low development and inadequate nutrition” remain significant challenges.
While a walking, talking, and synthetic brain is still a distant prospect, organoids have the potential to be a true lifesaver for people with neurological conditions. Just as other electronic brain interfaces, such as the Neuralink brain-machine interface (BMI), aim to improve the lives of people with neurological disorders, these organoids could potentially be implanted into living brain tissue to stimulate neuronal growth.
As the debate rages over whether the future will be built on human ingenuity or artificial intelligence, scientists are bringing these two worlds of intelligence closer together than ever before.
See also... Complete list of HTML/Javascript event handlers with examples List of required documents for a website in the Russian Federation |
