What capabilities are coming in GenAI’s near future…
Late last week Open AI internally released their “Stages of Artificial Intelligence”, and it’s worth digging into that a little to highlight the important achievements that are unlocked and stacked at each level.
Level 1 - Chatbots: Imagine talking to a computer that's indistinguishable from a human. This is the realm of advanced chatbots, which can hold conversations at a level that once defined artificial general intelligence (AGI) through the famous Turing Test.
Level 2 - Reasoners: Building on conversational skills, these AI systems can think at an expert human level. They can tackle complex questions using known information, bridging the gap between conversation and problem-solving.
Level 3 - Agents: These AI systems take it up a notch by executing complex tasks with near-perfect reliability. They can interpret instructions and complete multi-step challenges that require understanding and sequencing actions.
Level 4 - Innovators: Combining all previous abilities, these AI systems can generate new knowledge. They rapidly analyze data and iterate potential solutions, pushing the boundaries of what's known.
Level 5 - Organizations: The final stage involves coordinating multiple Stage 4 systems. Working together, these AI "organizations" produce results greater than the sum of their parts, potentially revolutionizing entire industries.
The most interesting part to the discussion of these levels of capabilities is that on the current development trajectory, we’re only 6-10 years away from achieving them all. Will we see a stalling out at certain levels? What might cause that? How will business, society, or whole countries change in the face of these achievements?
All fascinating issues that, like it or not, we are all going to live through to find out the answers together…but what do you think?
