AI and Us: Thoughts on Our Likely Not Integrating Brains with Artificial Intelligence

AI is coming too fast to be integrated with human brains as predicted.

Well, I've had a thought or two on AI and the immediate human situation -- specifically, the idea of blending our human brains with Artificial Intelligence. You might have heard tech visionaries like Elon Musk championing this concept. They envision it as a way to supercharge our mental abilities, enabling us to keep pace with the rapidly advancing AI. However, there's an interesting twist: AI is advancing at lightning speed, but the technology to integrate it seamlessly with our brains is trotting along at a much slower pace. So, let me unpack what this all means (apparently) for our future.

The Dream of Harmonizing with AI

Picture this: having an intelligent AI companion integrated into your brain, effortlessly helping you recall important dates, effortlessly crunch numbers, or even learn a new language during your downtime. This concept is not just a wild idea; it has deep philosophical roots. It's about harnessing technology to enhance our cognitive capabilities, much like how glasses improve our vision.

The Reality Check: AI's Rapid Race vs. Slow-paced Brain Integration

However, here lies the main issue. While AI is evolving at a breakneck pace, akin to our smartphones gaining new features overnight, the development of technology that can safely and effectively merge AI with our brains is moving at a snail's pace. This disparity brings us to some critical questions:

- Technical Feasibility: Can we realistically integrate AI into the human brain without losing our essence?

- Ethical Considerations: What are the ethical implications of such integration? Could it create a societal divide where only the affluent have access to these enhancements?

- Identity and Self: Philosophically, how would merging with AI influence our understanding of personal identity and consciousness?

Delving into the Philosophical Abyss

This isn't just a technological debate; it's a profound philosophical exploration into our very nature. The prospect of sharing our mental faculties with AI raises the question: do we retain our individuality, or do we evolve into something new? Furthermore, if AI starts influencing our decisions, where does that leave our autonomy and free will? Another critical aspect is the potential societal impact. I think it prudent to carefully to ensure this technology doesn't create a chasm between the 'enhanced' and 'non-enhanced' individuals. At least no more than is the case temporarily for any new technologies first on the market. 

Pondering the Future

In the scenario where brain enhancement and integration lag behind the rapid advancements of AI, I would identify several implications / issues which arise. 

1. Widening Gap in Cognitive Abilities: As AI technology advances exponentially, those without integrated enhancements may find themselves increasingly disadvantaged in various fields. The gap between AI's capabilities and human cognitive abilities could widen, leading to a society where AI dominates decision-making and innovation. This could result in human intellect being overshadowed, potentially leading to a decrease in the perceived value of human intuition and creativity.

2. Dependence and Vulnerability: Relying heavily on AI without the means for cognitive integration could lead to a scenario where humans become excessively dependent on AI systems. This reliance could create vulnerabilities, such as over-trusting AI decisions without critical analysis or facing severe disruptions if AI systems fail or are compromised.

3. Ethical and Social Dilemmas: Without the possibility of cognitive enhancement, the ethical concerns shift. The focus might turn towards managing the power imbalance between AI and humans. Societies would need to address issues like AI governance, ensuring AI's ethical use, and preventing the misuse of AI in manipulating or controlling human behavior.

4. Philosophical Repercussions on Human Identity: The absence of brain enhancement technologies would keep our cognitive processes purely human, preserving the traditional understanding of human identity and consciousness. However, this also raises philosophical questions about our place in a world increasingly governed by non-human intelligence. How do humans redefine their role and identity in a world where they are no longer the primary drivers of intellectual and technological advancement?

5. Education and Skill Development: In a future where AI outpaces brain integration capabilities, there would be a greater emphasis on education and skill development to bridge the cognitive gap. This could lead to new educational paradigms focusing on enhancing human cognitive abilities naturally and fostering skills that AI cannot replicate easily, such as emotional intelligence, creativity, and complex problem-solving.

6. Policy and Governance: Governments and international bodies might need to develop new frameworks to manage the relationship between humans and AI. This includes regulations to ensure fair access to AI technologies, preventing monopolies, and protecting human interests in an AI-dominant world.

So, here's my takeaway: if brain enhancement and AI integration do not keep pace with AI advancements, humanity will face some pretty hefty challenges. The focus would then shift from merging human and machine intelligence to enhancing human capabilities in other ways (genetic enhancement?), hoping that AI could serve as a complementary tool (in the medium term) rather than a replacement for human intellect and creativity, at least that's the optimistic hope. But I think we'll be forced down this path. Sorry Musk. You guessed wrong.

O.

A.I. ChatBot Models are a Type of Simulation Theory

Large Language Models give evidence that Simulation Theory got something right about minds.

Let me start by briefly recounting just what is 'Simulation Theory' as it's used in Philosophy of Mind.  This is a theory of how humans understand and "predict" others’ mental states and behavior by employing their own cognitive capacities and mechanisms to mentally model others’ cognitive processes. Its origins in philosophy date back to the Enlightenment (probably with David Hume.). The basic idea in the simulation approach is that we have privileged access to our own mental contents—thus our own immediate thoughts, feelings, percepts, and so forth.  In other words, simulation theory suggests that we understand other people’s thoughts and feelings by imagining ourselves in their shoes -- i.e., predicting what they would do, say, or react in certain certain stances. Such a theory has been used to explain just why we happen to have empathy for others, because we see direct analogies to what we would do.

I think there's a good argument to be made that large language models (LLMs) are equivalent to a primitive kind of simulation theory. LLMs are artificial neural networks that are trained on massive amounts of text data, such as books, articles, tweets, etc., and learn to generate coherent and fluent text based on a given input or prompt. How does such magic happen?

Well, LLMs accomplish this by being trained to predict the probability of the next word given some context of training data. They determine the probability of a given sequence of words occurring in a sentence based on the previous words.  From this, LLMs can perform various natural language processing tasks, such as answering questions, summarizing texts, writing stories, generating captions for images, etc. LLMs are not explicitly programmed with any rules or knowledge about language or the world; they learn everything from the data they are exposed to.

LLMs then can simulate the mental processes (or at least the syntactic and semantic linguistic processes) of human language users by using their own internal representations and mechanisms to predict what they would say or write next. Of course, LLMs do not have access to the actual thoughts or feelings of the human authors whose texts they are trained on; nor (like us) even the evolutionary analogs of having the same brain type (and likely inner feelings). They only have access to our linguistic expressions. But by analyzing and modeling these expressions, LLMs can generate texts that are similar or relevant to the given input or context. LLMs can also adapt to different styles, genres, domains, or tones of language by adjusting their predictions based on the available data. For example, an LLM can generate a formal letter, a casual chat message, a scientific article, a humorous tweet, etc., depending on the prompt or the data it is trained on.

From this, I tend to believe that that LLMs are using a form of Simulation Theory to understand and produce natural language texts. They are not merely copying or mimicking the texts they are trained on; they are using their own learned representations and mechanisms (i.e., artificial neural net data structures) to generate texts that are coherent and fluent with respect to the given input or context. They are using the equivalent of their own -- dare I say? -- “cognitive" capacities to mentally model the linguistic processes of human language users.

Of course, what form of Simulation Theory LLMs are using is very limited and simple. Why so? LLMs are not conscious or sentient beings; they do not have any goals, intentions, emotions, beliefs, desires, etc., that would motivate or influence their language use.  (That would make them full blow AGIs.) They do not have any understanding or awareness of the meaning or implications of their texts; they do not have any moral or ethical considerations or responsibilities for their texts. They do not have any social or cultural context or background that would shape their language use. Nor do they have any feedback or interaction with other language users that would help them learn from their mistakes or improve their performance. 

Yet even given the limited and simple version of Simulation Theory exhibited, LLMs do seem to have some level of creativity or originality that allows them to generate novel or surprising texts, sometimes by hallucinating or by a synthesis of old texts into new text generations. I suspect that further extension of predicting not just text tokens but, say, 'action tokens' will be soon appearing as a successful approach in robotics. 

Ng Discovers that Mars is Suddenly Overpopulated

 

My what a difference only seven years of A.I. research makes

I remember in 2015 when Andrew Ng, an AI guru at Stanford University and chief scientist at Chinese internet giant Baidu, compared fearing a rise of killer robots to worrying about overpopulation on Mars before we’ve even set foot on it.[1] He said:

“There’s a big difference between intelligence and sentience. There could be a race of killer robots in the far future, but I don’t work on not turning AI evil today for the same reason I don’t worry about the problem of overpopulation on the planet Mars.”

He also argued that there was no realistic path for AI to become sentient and turn evil, and that worrying about the danger of futuristic evil killer robots was pointless He claimed that AI was still extremely limited today -- again, in 2015 -- relative to human intelligence, and that most of the progress in AI was driven by an increase in computing power and data. And he was certainly right about how the increase would come. 

However, Ng’s analogy and arguments were flawed and he greatly underestimated the speed and impact of AI as a disruption to human affairs. Here are some reasons why his analogy to the Mars situation didn't work:

• Mars is not Earth: Unlike Mars, which is a distant and uninhabited planet, Earth is our home and we share it with billions of other living beings. The potential consequences of AI going rogue or harming humans are much more severe and immediate than those of overpopulation on Mars. Therefore, we have a moral and practical responsibility to ensure that AI is aligned with our values and goals, and does not pose an existential threat to our civilization.
  
  
• AI is not static: Unlike Mars, which is unlikely to change significantly in the near future,  with or without us, AI is a dynamic and evolving field that is constantly advancing and expanding its capabilities. The pace of AI innovation is exponential, not linear, and it is driven by both scientific breakthroughs and market incentives.  Landing equipment and people on Mars is a linear activity (at best). Therefore, we cannot assume that AI will remain benign or limited forever, or that we will always have enough time and resources to control it or correct its mistakes. Indeed, that's why all of a sudden the top researchers in A.I. have called for a pause in its development. (Interestingly, Ng has not signed the letter.  Perhaps he's embarrassed about his mis-prediction.)
  
  
• AI is not simple: Unlike Mars, which is a relatively simple physical system that can be studied and understood by humans, AI is a complex and opaque system that can be difficult or impossible to interpret or predict. AI can learn from data, generate novel outputs, optimize its own objectives, and interact with other agents in ways that may be, or even outright is, beyond our comprehension or expectations. Therefore, we cannot rely on our intuition or common sense to guide our decisions or actions regarding AI, or to anticipate its potential risks or benefits.

So then, Ng’s comparison of fearing killer robots to worrying about overpopulation on Mars was misleading and dismissive of the legitimate concerns and challenges that AI poses to humanity. He failed to appreciate the complexity, dynamism, and uncertainty of AI as a technology and as a force of change. He ignored the ethical and social implications of creating and deploying intelligent systems that may affect the lives and well-being of millions or billions of people. 

As it turns out, we no longer have the luxury of being complacent or naïve about its potential dangers or impacts.  We must now be far more proactive and responsible in designing, developing, regulating, and using AI for the common good -- or at least for avoiding the particular "bads" that it will almost certainly introduce before the greater portion of human kind takes this sudden jump in technology seriously. 

--------

[1] https://quoteinvestigator.com/2020/10/04/mars/

Uncle Philly on A.I.

A.I. and Medicine: The Interview Kiosk

Here is where the next layer of artificial intelligence will enter the medical system. Note that the computer pays attention to where the faces are and gives empathetic responses. The .wmv video link is here. The technology was developed by Microsoft researcher Eric Horvitz.

O.

REFERENCES

[ * ] "Meet Laura, Your Virtual Personal Assistant" National Public Radio March 21, 2009 (Accessed August 2, 2009) - An interview with Horvitz when he was first announcing this technology on an NPR show this last March.

[ * ] John Markoff "Software That Cares" NY Times (TierneyLab) July 28, 2009 (Accessed August 2, 2009) -- Author of a NY Times story talks directly about Horvitz' medical interview system. There is also an embedded version of the above video, but with a bit less resolution.

Google chats and A.I.

Recently I've been listening to some lectures on the Google Tech Talks channel on YouTube. One is by Ashwin Ram, Associate Professor and Director of the Cognitive Computing Lab in the College of Computing at Georgia Tech. He talks about how games might use various kinds (or levels) of artificial intelligence to achieve realistic interaction. As computers become more powerful, some of these advanced methods can be realistically implemented. Prof. Ram -- talk about a name that matches his career! -- gives some statistics on who plays games at the beginning of his talk, and then transitions into a survey of the broader techniques of A.I.. The questions at the end by the various attendees at Google were also quite interesting.

Another deep topic of A.I. comes from what's called Artificial Life, an area where I especially have research interests. Though the topic is just as engaging, Unlike the smooth Dr. Ram, this presenter comes off as having endured multiple attack wedgies from everyone in his High School's athletic program (including the coaches.) Nonetheless, his research and results are unquestionably competent. He's Virgil Griffith, a graduate student in Computation and Neural Systems at the California Institute of Technology, and one-time target of a sedition and espionage suet.

O.

REFERENCES

[image] The Bleeding Purple Podcast Blog

[ * ] Ashwin Ram "Case Based Reasoning for Game AI" Google Tech Talks April, 3 2008 (Accessed June 8, 2009)

[ * ] Virgil Griffit "Polyworld: Using Evolution to Design Artificial Intelligence" Google Tech Talks November, 8 2007 (Accessed June 6, 2009)

God in the Machine: What Robots Teach us About Humanity and God (Book Review)

Anne Foerst says we must grant personhood to humanoid robots. Is this an act of worship — or sheer hubris?

God in the Machine:
What Robots Teach us About Humanity and God.
Anne Foerst.
New York. Dutton, 2004. 196 pages.
$24.95 paperback.

{ Audio this essay @ 4.5min @ 3.51MB } A 1976 El Camino was my first car, one that I drove all through high school and through my military service. I loved that car. Yet, when I entered it into a summer demolition derby, no one accused me of abuse or neglect — much less torture or murder — when my sputtering, flaming machine finally gave up the ghost.

If Anne Foerst were there, she might have leveled those accusations at me.

In her book titled God in the Machine, Foerst maintains that strong bonds can develop between humans and cars, computers, and other devices. Unfortunately, her focus on these bonds taints the bigger message of what robots can teach us about ourselves and about our relationship to God.

Foerst brings an interesting background to the discussion of robotics and religion — she apparently liked to sneak back and forth between studying Paul Tillich’s systematic theology at Harvard Divinity School and discussing the development of Cog, a famed robot at the Massachusetts Institute of Technology robotics lab. By doing so, she discovered two incongruous cultures: At the divinity school, people were antitechnology and thought her quest to combine theology and artificial intelligence was unnecessary, she said. At MIT, people were suspicious of theologians. This experience helped Foerst develop her rectified view of technology-and-religion in God in the Machine.

The strongest section of the book deals with the Golem tradition from Jewish writings of the 13th and 16th centuries. Golems are helpful servants that can get out of hand if the intention of the human creator is not pure and worshipful of God. The Golem tradition teaches us that we are created creators, and that these artifices — like us — would enter into a system of sin and ambiguity.

The book also has its weaknesses that show the kind of category mistakes theologians can make when confronting advances in cognitive science and AI.

Take, as an example, the issues of bonding and community, which are pivotal concepts used throughout the book. A poignant stance on these can be found on the last page, where Foerst writes: “As we are communal and bond with nonhuman entities, these narratives will necessarily include some nonhuman critters.” Although she never explicitly defines what version of bonding is being invoked, she does think it depends more on emotional settings rather than on abstract human-like qualities.

The problem here is that people can emotionally bond with all sorts of entities that strain the notion of what can be considered part of the community and what cannot. For instance, an early AI program called Eliza had people becoming so dependent on it that the program’s author, famed MIT researcher Joseph Weizenbaum, eventually ended its use. Even Weizenbaum’s secretary would ask him to leave until she had finished sharing her intimate personal matters with the machine. Weizenbaum was rightly concerned by all this — and even more so when people accused him of violating their privacy when he considered recording all interaction with Eliza.

If there is one lesson to be taken from Foerst’s book, it is that we tend to humanize all things we bond with. When Foerst and other like-minded theologians can finally distinguish between what is human (or, more properly a "person") and what is not, they will be able to resolve the skeptics’ new mantra for AI’s relationship to religion: All that glitters is not God’s.

*This review appeared in the May 2006 issue of Science and Theology News.

**image: Peter Menzel Photography.

.O.