If somebody were to give you a timed future trajectory of AI, most likely you would not believe it. Take a step backward, when OpenAI was founded in Dec 2015 did Sam Altman have any idea of AI’s growth, or what it would turn out to be. GPT-1 came into existence in June 2018, and yet no one knew what it would finally do, and in which form it would bring AI to the world, if it did. Most of us, who were in the know of things, never believed that it would ever be a commonplace technology as it is today. Making GPT 3.5 conversational did the magic.
Whether it was planned in this manner or it just clicked, has not ever been public. But having tasted blood, Sam Altman, believed himself to be the tech messiah of mankind. The progress since the onset of ChatGPT has been mesmerising to say the least, with product after other falling out of the global AI closet. The AI ruling elite has slowly been formed, with the well-known risk of the outliers. When Sam Altman toured the world as the self-styled brand ambassador of the AI, the tech, business and funding model accepted without a murmur, was what professed, projected and many a times philosophised.
The DeepSeek model released this January, put both the AI compute and business model on its head. Singularity may come or not, but singularity of thought, on such a complex, all-encompassing topic, full of only moving parts is unacceptable. OpenAI’s CEO made headlines in late 2024 stating that AGI would arrive as early as 2025. He asserted it’s “basically an engineering problem” and the path to AGI “was not very clear.” There were not many takers of this prophecy and six months through in 2025, we don’t have an inkling of it. When the euphoria of OpenAI was at its peak, Satya Nadella once said that though Google has not been a pioneer in many technologies, but they dominated it later.
Sam Altman has done it again. He says AI has already gone past the event horizon but no worries since AGI and ASI will be a gentle singularity. AGI is human intellect, and ASI is beyond human intellect, superior to humans. In fact it is unknown whether we will reach AGI and hence equally fluid timelines. ASI is even beyond the pale. Latest polls suggest that AI experts generally believe that they will reach AGI in the year 2040. Sam’s recent blog does not clarify whether AGI is happening by 2030 or 2035 or whether it is ASI instead of AGI. Are both being used interchangeably? Franklin D. Roosevelt wisely stated: “There are as many opinions as there are experts.”
UNDENIABLY, AI TRAJECTORY REMAINS THE HOTTEST TECH TOPIC OF OUR TIMES.
Sanjay Sahay
Have a nice evening.