AI: OpenAI brings 'AGI' positively down to earth. RTZ #417

...the five levels to AGI 'Is the Way', for now

The Bigger Picture, July 13, 2024

In this Sunday’s Bigger Picture, I’d like to discuss how OpenAI is trying to bring the nebulous topic of timing to ‘AGI’ (aka Artificial General Intelligence), down to more concrete, and hopefully comprehensible steps by mainstream audiences full of Fear of AI, barely in the second year of this AI Tech Wave.

It’s a smart strategy, and likely resets the tactical stage for most LLM AI companies small and large going into the second half of 2024. As the main character in the Star Wars spinoff ‘The Mandalorian’ likes to say, ‘This is the Way’ for the AI industry for now. Let me explain.

As I described in yesterday’s weekly summary:

“OpenAI is cleverly redefining the raging ‘time to AGI’ debate, laying out 5 levels to AGI. The Levels go from L1 ‘Chatbots’ (here now), to L2 ‘Reasoners’, L3 ‘Agents’, L4 ‘Innovators’ and L5 ‘Organizations’. No timing is specified, although as Ben’s Bites notes, OpenAI is intimating it is likely at the threshold of L2 presumably with GPT-5 later this year.”

“This reframing will now likely be responded to by all other LLM AI companies, from Google to Meta to Anthropic and others. The 5 level rating system of course echoes the Government’s L5 rating system on self-driving cars (aka and akin to Tesla’s Full-Self Driving, FSD AI tech), a topic I’ve discussed in depth.”

“It also fits in with OpenAI Founder/CEO Sam Altman’s desire to transition to an ‘iterative’ approach to AI product and service innovation, and less a ‘Big Bang’ to AGI (Artificial General Intelligence), that has gripped the industry and regulators. Notable that ‘Agentic Workflows’ blend both L2 reasoning and L3 smart agents, and is a core focus for most AI companies already, as discussed here.”

Reframing AI & AGI from Negative to Positive.

OpenAI’s effort above is laudable as it brings the epheremeral existential fears around AGI down to earth. And reframes them in a more concrete set of positive steps. They may not all be in the sequence above, but directionally point the way.

The folks inventing the tech and now AI future have long been inspired by the Scifi interpretations over the years, sometimes decades before their time. And the existential fears of AI today are inspired by visions of a ‘Terminator Skynet’ future, amongst countless others.

Star Trek, from the original 1966 series on, has inspired the more optimistic visions driving the founders of countless technology companies to Datel. From Google to SpaceX to OpenAI today and beyond.

So many innovations of previous tech waves, from the PC, to the Internet, to Mobile and now AI, have been driven by trying to pull the future forward from Star Trek. And LLM/Generative AI is having a go at doing a little more tugging, as I recently described in some X/Twitter posts that are relevant in today’s discussion.

The Star Trek ‘Universal Translator’ Future, that I’ve discussed before, is getting even closer for example with the ‘AI Interpreter’ feature on Samsung’s new Buds 3 Pro wearable wireless headsets. It offers real-time translation to and from languages like Chinese, English, French, German, Hindi, Italian, Japanese, Korean, Polish, Portuguese, Spanish, Thai, and Vietnamese. For now. And Apple Air Pods Pro will likey follow soon and others.

To see the latest glimpse of that future, see this great video review by Flossy Carter on X/Twitter, with the AI Interpreter featured shown at the 13 minute mark. Note his enthusiasm and excitement. AI delight all the way. A ways to go of course to Star Trek ‘reality’, but a baby step in that direction. Lots of L2/L3 ‘Reasoning’ and ‘Agents’ eventually in just this application to come.

Next up I discussed on X/Twitter, is the Star Trek ‘Holodeck’, an imaginary, ‘magical’ place on the Starship Enterprise that converts a three dimensaional room into a real stage with characters from any human media, from books to movies.

A demonstration by AI luminary Andrej Karpathy (formerly of OpenAI/Tesla and others), shows how he uses LLM AI software and tools to convert characters from Pride and Prejudice into an animated movie. As he explains:

“I'm playing around with generative AI tools and stitching them together into visual stories. Here I took the first few sentences of Pride and Prejudice and made it into a video.”

Again, baby steps, but with exponentially scaling AI that I’ve discussed, all manifestly possible as mainstream products and services, not too far into the future. Lots of travel here across OpenAI’s AGI levels in just developing these capabilities for mainstream consumption globally.

Next up, I discuss the Star Trek ‘Tricorder’ and Communicator’ also being pulled into the present a bit faster than expected:

“Great threads on both by @LinusEkenstam. They’ve inspired the smartphones and early ‘AI pins’ in these very early days of the AI Tech Wave. But as I discuss with @ramahluwalia next up are the baby steps of AI ‘Reasoning’ and ‘Agents’ (Level 2 and Level 3 stages by @OpenAI)”

I’ve discussed Reasoning and Smart Agents as the next big things to scale exponentially with AI. Both of course are OpenAI’s Sam Altman 5 layer AGI roadmap discussed above.

And that’s why these collective forays into more optimistic Scifi scenarios from old TV shows and movies. To try and connect words like ‘AGI’, ‘Levels’, ‘AI Reasoning’, ‘AI Agents’ into ideas we’ve emotionally experienced in our imaginations of the possible future. (Here are some video shorts collections of my discussions around these and other AI terms and issues for those interested).

As noted earlier, there has been a raging debate for almost a decade on the timing to AGI, with extreme views on both sides, negative and positive. The answer as always in tech waves, is generally somewhere in the middle, over time. And it generally takes a lot MORE time than envisaged in the beginning.

I’ve long believed that it’s premature to fear the ‘existential’ downsides of the AI. Especially before we have any sense of mainstream AU applications of these rapidly evolving technologies from a bottom up perspective. Always driven by organic inventions and innovations by entrepreneurs, engineers, and eager investors at every end of the investment spectrum globally.

OpenAI’s five levels to AGI is one way to ‘The Way’. It reframes the path to AGI as an iterative series of steps. A journey, not a ‘Big Bang’ destination in one go.

It’s going to be a volatile, but eventually rewarding ride, even if it’s not at ‘Warp Speed’ from Star Trek. We’ll have to wait a bit longer to start positively pulling in that future. That is the AI Bigger Picture’ this Sunday. Stay tuned.

(NOTE: The discussions here are for information purposes only, and not meant as investment advice at any time. Thanks for joining us here)