Typically, a birthday is the anniversary to remember and rejoice special moments of love, affection, loyalty and many hues of emotions overflowing in one’s life. The previous week marked the first anniversary of launch of ChatGPT (GPT-3) on 30 Nov 2022 – the pathbreaking generative technology powered large language model (LLM) introduced in the public domain for people to explore for the first time. Around the time of its first anniversary, intense melodrama of boardroom coup at OpenAI and dramatic turn of events of its CEO being fired and rehired withing a week, left a bewildering note than a moment of jubilation of its astounding success. Worrisome news indicates that OpenAI staff’s alert to the board about an AI breakthrough – which could threaten humanity, could be one of the possible trigger points, besides lack of trust and communication.

Intensifying Artificial Intelligence arm race

Amid intensifying AI race and Big Giants’ surreptitious AI experiments to command artificial general intelligence (AGI), Google DeepMind has now launched Gemini - multimodal general AI models in varying sizes – Ultra, Pro and Nano. Google claims that Gemini Ultra (to be launched to power Bard Advanced in coming period) exceeds current state-of-the-art results on 30 of the 32 widely-used academic benchmarks used in LLM research and development. Having introduced generative pre-trained transformer (GPT), generative AI, LLM, deep learning, prompt engineering, reinforcement learning from human feedback (RLHF), massive multitask language understanding (MMLU), search generative experience (SGE) and many innocuous terms in lingua franca, the lexicon is bound to get further enriched with new AI-laced acronyms.

Beyond the linguistics and etymological exposition, the bigger question emerges –what are the perceptible impacts and more importantly, unanticipated influences unbeknownst to the humanity from the advent and unbridled advancement of AI, galloping toward AGI.

Convoluted design of democratization of AI

While touting about tremendous potential of AI to drive innovation and augment human abilities, we conveniently overlook on the basic requisites for ethos of democratization of AI to naturally prosper. Even if we celebrate phenomenal growth of ChatGPT attracting 180.5 million users in a short span of time, it is barely around 3.4% of total internet users in the world or 2.2% of population of the world. Very much like mainstream political systems with skewed set-up of democratic moorings, the power of AI technology is bound to remain concentrated in the hands of a few wealthy, resourceful and wily organizations and individuals – whose success lies in restricting the access, making its less affordable and obfuscating sharing of research-based knowledge with economically and technologically backward societies.

In recent years, regulators have been raising concerns about AI adding new vulnerabilities and unconventional risks, which are less obvious, hard to audit, and have limited explainability. Also, enough has been written about dangers of AI powered models exacerbating data privacy, security, model governance, and third-party related risks, besides copyrights infringements.  Curiously, other than prolonged reflections on AI risk management and governance and occasional toothless(?) regulatory guidelines (EU Artificial Intelligence Act being an exception), it appears to be an insipid affair globally. Regulatory formulations across the jurisdictions, at the best, sluggishly follow the technological advancements and remain hopelessly patchy in coverage.

Everything comes with a price: Is it worth it?

Technological inventions and advancements have been part of human endeavors, for prosperity and welfare of the humanity. More than 2 centuries after invention of steam engine and its full-scale adoption, people felt exhilarated with enhanced means of mobility and productivity gains as well as new level of prosperity in surrounding societies. With rapid expansion of railway network, a plethora of new job opportunities were created. People hardly felt so threatened about the machine stealing their job and making them redundant. For that matter, massive wave of computerization in late 20th century, even after sparking initial fear of job loss, never degraded staffs’ job profile rather transformed the quality of job content with automated workflows and productivity tools - ultimately adding to their proficiency and income levels. Does the same evolution path hold true for turbocharged and unstoppable game of AI? Either no one has an answer yet, or those having foresight and insight possibly would like to remain tight-lipped – much like Noah was silent before God knowing about the coming destruction of the world.

Human supremacy and inventiveness facing AI tsunami

Unrestrained proliferation of AI sets the alarm about concentration of technological power and its parochial applications, which makes wealthy being obnoxiously wealthier – well reflected in soaring market cap of Big Tech giants and personal networth of their rockstar founders and investors. Engrossed in mindless maneuverings in guise of innovation – unmindful of permanent injuries to the larger human society, it creates a new pattern of technology-powered racism. Remember, Sam Altman’s imperious scoff off on the idea of a young team from India with $10 million could build foundational model something like OpenAI.

The pages of history have enough lessons - while controlling the flows of technology and know-how, destroying the self-esteem and confidence of a society can be an early sign of colonialism. Of course, related perils of economic degradation and dependency as well as imposed socio-cultural supremacy can be uninvited outcomes. The same holds true for AI powered commercial ventures and global enterprises – intently infringing each part of our lives, even if the Doctrine of Discovery of Middle Ages appears incongruous at the face in the 21st century. While being open-minded and assimilative to lap technological innovations for augmentation of human abilities, it would be curious to see as at what point people turn the tide of AI tsunami by limiting its intrusive power and make it subservient to the ultimate human supremacy with all faculties of human inventiveness intact.

Image source: https://www.imf.org/en/Publications/fandd/issues/2023/12/Macroeconomics-of-artificial-intelligence-Brynjolfsson-Unger