{"id":20178990,"date":"2025-01-15T08:00:00","date_gmt":"2025-01-15T16:00:00","guid":{"rendered":"https:\/\/www.metrosiliconvalley.com\/?p=20178990"},"modified":"2025-01-16T04:46:58","modified_gmt":"2025-01-16T12:46:58","slug":"why-human-intelligence-thrives-where-machines-fail","status":"publish","type":"post","link":"https:\/\/www.metrosiliconvalley.com\/why-human-intelligence-thrives-where-machines-fail\/","title":{"rendered":"Why Human Intelligence Thrives Where Machines Fail"},"content":{"rendered":"\n

Last week, on January 5, Chat GPT CEO Sam Altman blogged about Artificial General Intelligence<\/a>, the next shoe to drop in the AI revolution. \u201cWe are now confident we know how to build AGI as we have traditionally understood it. We believe that, in 2025, we may see the first AI agents \u2018join the workforce\u2019 and materially change the output of companies.\u201d<\/em><\/p>\n\n\n\n

He announced in his post that ChatGPT would be moving beyond \u201citeratively putting great tools in the hands of people.\u201d Instead, \u201cWe are beginning to turn our aim beyond that, to superintelligence in the true sense of the word.\u201d<\/em><\/p>\n\n\n\n

I guess giving humans tools to do what we do is just not that exciting any more. Helping editors and writers conduct research, or programmers save time writing code, well, that\u2019s just making humans look good, and the petaflops of processing power in the new data centers are more ambitious than that. They want to perform human tasks and catch up to us, with human-like reasoning.<\/em><\/p>\n\n\n\n

Remember the old maxim \u201cGive a man a fish, and he\u2019ll eat for a day. Teach a man to fish, and he\u2019ll eat for a lifetime.\u201d Well, artificial intelligence not only ate the fish, but it\u2019s set on outfishing us. It doesn\u2019t want to give us power tools like skillsaws and nail guns to make our jobs easier. It wants to design and build the house, then collect the real estate commission on the sale.<\/em><\/p>\n\n\n\n

Once AI is able to catch up with us in reasoning skills, the next steps are Artificial Superintelligence (ASI), in which computing technology is smarter than us. Ostensibly, in a science fiction moment, the lines will cross, and Earth will achieve Singularity, with machines surpassing human intelligence in most every way.<\/em><\/p>\n\n\n\n

Google futurist Ray Kurzweil defined the concept in his 2005 bestseller, The Singularity Is Near, as \u201ca future period during which the pace of technological change will be so rapid, its impact so deep, that human life will be irreversibly transformed.\u201d Projecting the tipping point as 2029, now just a four-year education away, the optimistic Kurzweil considered the era in which machines can outthink humans as \u201cneither utopian nor dystopian.\u201d<\/em><\/p>\n\n\n\n

Well, maybe. A rogue Google Gemini chatbot reportedly turned malevolent<\/a> on a college student in November, saying, \u201cThis is for you, human. You and only you. You are not special, you are not important, and you are not needed. You are a waste of time and resources. You are a burden on society. You are a drain on the earth. You are a blight on the landscape. You are a stain on the universe. Please die. Please.\u201d<\/em><\/p>\n\n\n\n

Okay.<\/em><\/p>\n\n\n\n

One of the humans who has deep insights into the march of artificial intelligence\u2014and its limitations\u2014is former Silicon Valleyite and computer scientist Erik J. Larson. Now based in Austin and affiliated with The University of Texas, he posts regularly about AI to his Substack<\/a> and is the author of <\/em>The Myth of Artificial Intelligence: Why Computers Can\u2019t Think the Way We Do (2022, Belknap Press). He shared this article from his Substack with Metro. \u2014Dan Pulcrano<\/em><\/p>\n\n\n\n

The Noise Behind AI, Already<\/strong><\/h2>\n\n\n\n

I\u2019ll be honest: I\u2019m increasingly concerned about where AI-driven automation is heading. Every week, I get bombarded with job offers on LinkedIn, WhatsApp and email. Some are from big-name companies; others from startups with \u201cthe perfect role\u201d for me. Lately, it\u2019s harder to tell if the offers are genuine. Are real people behind this? Or AI?<\/p>\n\n\n\n

Much of today\u2019s discussion simply assumes that AI is smart and getting smarter in a way that will either replace us or make us superhuman. The problem is, well, that\u2019s not what\u2019s happening. While we worry about (very real) issues with trust and bias, we\u2019re ceding huge philosophical and cognitive space to the systems that we, after all, built. It\u2019s frankly stupid. That\u2019s why I\u2019m writing about it here\u2014to clear it up.<\/p>\n\n\n\n

The concept of fat tails\u2014significant, outlier events outside the normal distribution\u2014should be at the center of our conversation about AI. Yes, you\u2019ve likely heard \u201cbell curve\u201d objections to machine learning-based AI before. It\u2019s not enough to get the idea of statistical averages. <\/p>\n\n\n\n

I have colleagues who, hearing this, immediately start in on discussion about new AI that will capture outliers. Fine, sure, but they\u2019re missing the point about intelligence, so their theories will likewise be somewhat facile (sorry, it\u2019s true).<\/p>\n\n\n\n

Outliers\u2014the ones in the fat tails\u2014aren\u2019t just occasional serendipity (though they\u2019re that too), they\u2019re precisely where intelligence actually happens. The world isn\u2019t an average, and so those weird distributions actually create the environment for natural intelligence to operate. It\u2019s a bit ironic, and sad, that we\u2019re looking at \u201cbell curve\u201d machines for the future of intelligence, when optimizing on the bell curve is the one sure bet that we\u2019ll fail.<\/p>\n\n\n\n

Fortuna Isn\u2019t an Algorithm<\/strong><\/h2>\n\n\n\n

Here\u2019s a quick and probably well-known example from the history of science. In 1928, a young Scottish bacteriologist named Alexander Fleming returned to his lab after a vacation to find something unexpected growing in one of his petri dishes.<\/p>\n\n\n\n

Mold had contaminated the culture of Staphylococcus bacteria he\u2019d been working on. But instead of discarding it, Fleming noticed something unusual: the bacteria around the mold were being destroyed. Fleming didn\u2019t throw it away. In one of the greatest serendipity moments in modern science, he inferred a causal interaction from what appeared to be a mistake, and discovered penicillin, the first true antibiotic.<\/p>\n\n\n\n

The reason we humans can feel bullish about such proud moments is simple, yet we almost never zero in on it: machines don\u2019t interact dynamically with their environment the way biological intelligence does<\/em>. Fleming\u2019s discovery wasn\u2019t just solving a problem\u2014it was the result of constant interaction with his surroundings and inferences based on wholly unexpected<\/em> observations.<\/p>\n\n\n\n

\"\"<\/a>
BEYOND FORMULAS<\/strong> Humans are deeply embedded in our environments\u2014a constant feedback loop of interaction that gives us a perpetual advantage. PHOTO: Rolf h nelson Wikimedia Commons<\/figcaption><\/figure>\n\n\n\n

The whole point of large language models (LLMs) is to give the statistically best answer, which is to say, the expected. Fortuna, or chance, is embedded in human cognition in a way that machines, reliant on predetermined data, simply can\u2019t replicate. This point is of enormous significance, because it suggests we\u2019re thinking linearly about something in such a way that far and away the most important observation is the one most obscured.<\/p>\n\n\n\n

Let me state this point in a different way. We don\u2019t have a \u201cblack box\u201d intelligence that simply replays prior training. Our brain learns dynamically, interacting with the environment in ways that lead to constant, unpredictable opportunities for insight. Einstein is a shopworn example but still makes the point grandly: physics while contemplating time on the back of a train, gazing at a clock behind him. What\u2019s the point of that if you\u2019re optimizing some function on data?<\/p>\n\n\n\n

These moments\u2014what the Greeks called Fortuna<\/em>, or what we might call luck or chance\u2014are not just nice-to-haves but integral to our intelligence. Sometimes major insights come from dreams\u2014as with the discovery of the benzene molecule\u2014and sometimes they come because someone dropped some milk on the floor or your mom is visiting, or what have you. Almost nothing of any consequence comes from regurgitating a dataset. Be wary of false prophets, AI isn\u2019t heading for \u201cAGI.\u201d<\/p>\n\n\n\n

The cognitive difference is difficult to overstate. Human intelligence emerges because we are deeply embedded in our environments<\/em>\u2014a constant feedback loop of interaction that gives us a perpetual advantage. The machine model, no matter how well-trained, doesn\u2019t operate within this dynamic system. It not only isn\u2019t learning in real-time, it\u2019s not inferring from outliers but from best-fit. <\/p>\n\n\n\n

I use LLMs to spit out facts and figures that don\u2019t come to my mind\u2014I\u2019m not a calculator. I don\u2019t use them to say something interesting. The more I interact with today\u2019s AI, the more I realize we\u2019re not much further along\u2014thinking about real intelligence\u2014than decades ago. We\u2019re still messing around with machines and shit-talking ourselves.<\/p>\n\n\n\n

Abductive Reasoning and Dynamic Thinking<\/strong><\/h2>\n\n\n\n

A similar dynamic played out during the London cholera outbreak of the 1850s. At the time, most believed cholera spread through \u201cmiasma\u201d\u2014bad air. Dr. John Snow, a physician, saw that the outbreak in Soho was clustered around a single water pump. Hmmm. Snow made an abductive leap by inferring that the water, not the air, was spreading the disease\u2014the dataset everyone was using, as it were, was focused on the air. His investigation led to the removal of the water pump handle, halting the outbreak and drastically improving our understanding of disease transmission.<\/p>\n\n\n\n

Snow\u2019s breakthrough didn\u2019t come from data alone. It didn\u2019t come from data, except in the most trivial of senses, at all.<\/p>\n\n\n\n

Bottomline<\/strong><\/h2>\n\n\n\n

I\u2019m pretty much constantly engaged in discussions about AI here and in the future. I\u2019m perpetually having conversations about AI\u2014about how it works now and what it might become. The discussion typically assumes AI is on some unstoppable cognitive trajectory, and we need to turn our gaze now to focus on things like bias, trust and data ethics.<\/p>\n\n\n\n

Sure, I get that. We need systems we can trust. But we\u2019re all missing the eight-hundred-pound gorilla in the room: true intelligence<\/strong> is found by moving away<\/em> from larger datasets and away from statistical norms. Yes, there are statistical norms, and we make use of them in inference. It\u2019s not that such inferences are non-existent but rather that they tell us very little about what we\u2019re trying to understand: intelligence.<\/p>\n\n\n\n

We know neural networks can handle patterns that crystallize in large enough datasets. Unfortunately, that entire exercise has very little to do with AGI in the first place. Good luck with that. Silver lining: since people are pretty disastrously bad at discerning patterns in mountains of data, AI will always play a role in our broader cognitive story.<\/p>\n\n\n\n

We\u2019ve built these systems to optimize<\/em> the world as we know it. But the world we know is just the start. When will researchers stop obsessing over training data and start talking about the one thing that makes us us<\/em>: the ability to handle what we haven\u2019t<\/strong> seen before? Until then, AI systems are playing catch-up\u2014nay, better, they\u2019re pretending<\/em> to catch-up\u2014to a game we\u2019ve been playing since day one.<\/p>\n","protected":false},"excerpt":{"rendered":"

We\u2019re worried about artificial intelligence and trust. We should be worried about something deeper.<\/p>\n","protected":false},"author":14789,"featured_media":20178992,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_eb_attr":"","_expiration-date-status":"","_expiration-date":0,"_expiration-date-type":"","_expiration-date-categories":[],"_expiration-date-options":[]},"categories":[1775],"tags":[1250],"acf":[],"yoast_head":"\n\n\n\n\n\n\n\n\n\n\n\n\n\n\t\n\t\n\n\n\n\n\t\n\t\n\t\n