The road towards artificial Intelligence that truly thinks with human-level consciousness and understanding is a marathon and no sprint.
The question has captivated a few scientists and industry because humans would be left stunned questioning their relevance in the world, and the reality that humans could arguably be in some potential danger.
The truth remains that no one knows what happens at that precipice. The point where AI has officially surpassed human-level intelligence, not only in outflow but in human-level consciousness and understanding.
How close are we to achieving this goal?
Progress has been made on the path toward AI that thinks independently.
One of the most promising avenues toward thinking AI lies in Large Language Models (LLMs) and their complex algorithms trained with massive datasets of text and code to allow them to generate human-quality text, translate language, or write different kinds of creative content.
Last October, Google unveiled PaLM 2 a Language model that demonstrated significant progress in various areas which include reasoning and factual language understanding.
In February 2024, OpenAI introduced GPT 4 which brought an interesting refreshed era with over 100 parameters, showcasing increased scale in LLM capabilities.
These were impressive milestones when looking at AI leaps since their public inception post-COVID-19.
As impressive as AI growth has been, experts warn against mistaking these advancements for true sentience or consciousness. LLMs still remain at a fundamental level statistical machines, due to their excellent pattern recognition and mimicry.
They still lack the ability to truly understand the meaning behind the words they generate. This was supported by the associates in the Advancement of Artificial Intelligence in April 2024 where researchers found that even the most advanced LLMs battled with tasks that required authentic comprehension and reasoning.
The huddle towards conscious AI
A major huddle that stands in the way of conscious AI could lie in the manner in which AI processes information.
One research report wrote humans not only process information but also interact with the world through our bodies, this bodied experience shapes our understanding and learning.
Current robots lack the sensory capability and dexterity to interact with the environment in a way that could contribute towards true intelligence.
The right to think
Another huddle in the road toward sentient AI could be rights, legislation, and potential dangers.
Researchers have called for a cautious and responsible approach to developing advanced AI. Elon Musk was also in the group emphasizing the need for transparency, safety and alignment with human values.
How far till D-day?
We are not there yet.
This is a fact, as in 2024, no organization has been able to deliver sentient AI.
This means the race is still on as rapid advancement in large language models and ongoing research offer a glimpse into the future possibility of sentient creations.
The road to a goal does require collaboration between researchers, ethicists, and policymakers in order to develop a responsible path towards an era unimagined.
Could we develop AI that understands the world and its place within it, could we ensure that AI development is aligned with human values?
These are some of the imperative questions that need answers to bring forward responsible AI that could be used for good.
While the quest towards conscious AI is an obvious marathon, significant progress has been made in recent years. The coming few years will be crucial in shaping the future of AI and its overall impact on humanity.