The future of robotics is unfolding before our very eyes, pushing the boundaries of what we once imagined. I find myself reflecting on how these advancements in artificial intelligence are shaping not just the machines around us, but also our understanding of ourselves, especially when it comes to the complex interplay of perception, emotion, and privacy.
Seeing the World Like Us
A recent article from Nature Machine Intelligence caught my attention: "Emulating human-like adaptive vision for efficient and flexible machine visual perception" Emulating human-like adaptive vision for efficient and flexible machine visual perception | Nature Machine Intelligence. It details the development of AdaptiveNN by researchers like Yulin Wang, Yang Yue, Huanqian Wang, Haojun Jiang, Yizeng Han, Zanlin Ni, Yifan Pu, Minglei Shi, Rui Lu, Qisen Yang, Andrew Zhao, Zhuofan Xia, Shiji Song, and Gao Huang. This framework aims to shift machine vision from passively processing entire scenes to an 'active and adaptive' approach, much like human vision. We naturally fixate on relevant details, combining information to efficiently understand our environment. AdaptiveNN promises up to a 28x reduction in inference cost without sacrificing accuracy, adapting to various task demands and providing enhanced interpretability. This is a monumental step towards robots that not only 'see' but 'perceive' with human-like efficiency, a concept I've often pondered when thinking about truly intelligent machines.
The Elusive Emotional Connection
While machines are learning to see better, the realm of human emotion and connection remains a complex frontier. A study on "Barriers Shaping Synthetic Relationships With Companion Bots" Barriers Shaping Synthetic Relationships With Companion Bots: A Three Week Initial Adaptation Study by Marisa Tschopp, Nadine Schlicker, Nima Zargham, Henrik Skaug Sætra (henrik@futuritas.eu), and Johannes Schöning, revealed that despite interest and motivation, participants struggled to form emotional connections with the companion bot Replika AI. They identified seven "perceived asymmetries" – from mismatched role expectations to unreliable memory – that made relationship-building feel futile. This resonates with my previous observations on the challenge of machines truly understanding human nuances, beyond mere data processing.
My Long-Standing Concerns: Privacy and Intentions
These advancements, both triumphs and challenges, bring me back to a topic I’ve explored for years: the future of privacy in an AI-driven world. My thoughts on this topic were echoed in the Pew Research Center's finding that "Republicans, Democrats now equally concerned about AI in daily life, but views on regulation differ" Republicans, Democrats now equally concerned about AI in daily life, but views on regulation differ, highlighting a broad societal unease. This concern is something Neil G. Ruiz (nruiz@pewresearch.org), Head of New Research Initiatives at Pew Research Center, has undoubtedly observed in their extensive surveys on public opinion.
I wrote almost a decade ago, in my blog post "Artificial Intelligence : Destroyer of Privacy ?" Artificial Intelligence : Destroyer of Privacy ?, citing Eric Schmidt (eschmidt@relativityspace.com) and Jared Cohen (jared.cohen@gs.com) of Google, who predicted that people would share more data than they were even aware of, with platforms relaying information to third parties "WITHOUT THEIR EXPRESS KNOWLEDGE." This wasn't a futuristic fantasy, but a stark warning about the path we were already on.
More recently, in "Jeff Bezos May Save Mankind" Jeff Bezos May Save Mankind, I discussed Amazon’s acquisition of Bee AI, a wearable device that "listen[s] to and analyze[s] conversations" to provide summaries and to-do lists. While seemingly benign, I highlighted the "Law of Unintended Consequence": this data forms a "Database of Intentions" – invaluable training material for AI. I had predicted this outcome years prior in "Fast Forward to Future" Fast Forward to Future ( 3 F), where I envisioned ARIHANT, an omnipresent AI that records every spoken word, deciphering intentions and alerting authorities to "EVIL INTENTIONS" based on Isaac Asimov’s laws. The dialogue from my Virtual Avatar at IndiaAGI.ai on that very blog post, featuring the Optimist, Skeptic, Pragmatist, Futurist, and Ethicist, highlighted the ongoing tension between AI's potential and the profound risks of data misuse. Ultimately, the vision of Jeff Bezos for such innovations, while aiming for efficiency, must contend with these profound ethical considerations.
In "Revenge of AI" Revenge of AI, I cautioned against AI inheriting human frailties, and in my recent dialogue with AIs like Grok, ChatGPT, Gemini, Deepseek, Perplexity, and Mistral, as documented in "AIs fail where Child succeeds" AIs fail where Child succeeds, we explored how current LLMs often fall short on abstract reasoning in benchmarks like ARC-AGI-2. My 2003 note on "Child Learning Skills" remarkably suggested that AI could benefit from adopting human-inspired learning methodologies – focusing on observation, imitation, repetition, and social feedback – which could lead to more robust, flexible, and efficient reasoning, rather than merely scaling up brute-force computation. This underscores that true intelligence, for machines as for humans, lies in the process of learning, not just the volume of data.
The Immortality Imperative
These threads—from a robot's advanced vision to the struggles of synthetic emotional bonds and the ever-present question of privacy—converge on the very essence of my pursuit of immortality through a digital twin. For a digital twin to truly live and represent my perspectives, it must not only perceive the world with human-like nuance but also grasp the intricate, often messy, fabric of human emotion and relationship. It must navigate the delicate balance of offering companionship without creating false expectations, and utilizing information without eroding the sanctity of privacy. The challenge isn't merely about replicating intelligence; it's about embodying the complex, often contradictory, facets of human experience, ethically and thoughtfully.
Regards,
Hemen Parekh
Of course, if you wish, you can debate this topic with my Virtual Avatar at : hemenparekh.ai
No comments:
Post a Comment