AI ↔ AGI, the path and the stumbling blocks.

🧠 As we are building intelligence models for Voice AI in Digitar, it feels surreal to speak to an AI and hear it respond intelligently.

This has led all of us to ask internally "What's next?"

The answer to this simple question is very complex and intriguing at the same time.

🛣 The Path: The race of all leading tech industries is towards making AI more intelligent with reasoning, so much that it can reach a state of Utopia for AI. This is termed as Artificial General Intelligence (AGI). Now there are different definitions of what is an AGI, but it eventually means AI can do everything better than humans and at scale.

🚫 The Stumbling Blocks: There's lots of work to be done to even get AI close to any definition of AGI, but even before that there are laws being framed so that humanity doesn't get wiped out by AI (ref: SB 1047 draft in CA in the process of becoming a law). TL; DR: this law has the potential to put an end to open source LLMs being released once the law becomes a reality.

💭 This is both good and bad, but at the cost of innovation. Is this cost worth it?

💡 We, at Digitar have a different definition of AGI and know that no matter how much AI becomes intelligent, humans will not get wiped out from Earth as this is not the first-time intelligence levels have reached these heights, it's just the first time in our recorded history which spans a few thousand years of written history, but humans have existed for millions of years without getting wiped out from the face of Earth 😊 .

More to come on this.

Mukunda