@kallaway You know J.A.R.V.I.S from Ironman?
What if I told you, we’re actually way closer to building it than people realize.
To make J.A.R.V.I.S, all you need is an AI system that can do 5 things.
First, you have to be able to communicate with it.
It needs to hear you clearly and speak back fluidly.
But we already have AI tools that can do this really well.
Eleven Labs and Hume AI are world class voice models that are already at J.A.R.V.I.S’s level.
Second, J.A.R.V.I.S needs to be able to see things and visually understand what’s going on.
It can’t just have ears, it also, needs to have eyes.
And those eyes wouldn’t be through a single humanoid, it would be through a series of cameras all over your space.
But again, we already have world action models that are getting pretty good at seeing and understanding.
But the third piece is the biggest one…
In order for J.A.R.V.I.S to work, you need one central AI brain that can essentially hop around freely to different robotics and smart devices throughout your space.
It’s kinda like seamlessly switching sonos speakers from room to room, but on steroids.
Today our smart devices are actually pretty dumb.
So to have a legit working version of J.A.R.V.I.S, we’d need an interconnected system of devices and machines that is smart enough to activate as you maneuver around it.
So that’s the brain and body, still a ways to go.
Part 4 and 5 is really what makes J.A.R.V.I.S, J.A.R.V.I.S.
And that’s the customizable personality and the proactive action system.
On the personality front, we really just need better memory and emotional expression.
But this is also coming and the progress in memory is obvious in the latest versions of ChatGPT.
The proactive action system is having J.A.R.V.I.S assess and be able to do stuff without you asking.
This is essentially an AI agent system.
Ironically the biggest problem holding back J.A.R.V.I.S today is not that our AI isn’t smart enough…
It’s that our cameras and robots are too dumb….for now.
Follow Kane K. for more videos like this.
Tried something new with this one, lmk what you thought!
#ai #artificialintelligence #tech #technology #ironman #jarvis #tonystark #robot
Explaining the tech behind J.A.R.V.I.S.from @kallaway
[01]
Overview
The creator breaks down the five essential components needed to build a real-life version of J.A.R.V.I.S. from Iron Man. He explains each step: (1) communication via advanced voice models, (2) visual understanding through cameras, (3) integration with a central AI brain controlling smart devices, (4) customizable personality through memory and emotion, and (5) a proactive action system. He concludes that the primary obstacle isn't the AI's intelligence but the lack of sophisticated, interconnected hardware.
A fast-paced, blurry, purple graphic animation of the J.A.R.V.I.S. interface, creating a dynamic and high-tech visual opening.
[07]
Hook and payoff
The promise the opener makes, and how the video pays it back
Hook
A 'What If' question that connects a beloved fictional concept to a present-day reality, creating a curiosity gap by claiming it's more achievable than the audience thinks.
Payoff
A counterintuitive reveal that the biggest obstacle preventing the concept's realization is not what the audience was led to expect, creating a memorable and shareable twist.
[08]
Narrative framework
The structural pattern this video runs on
Framework
The Deconstruction & Diagnosis Framework
To take a complex, aspirational concept and break it down into a numbered list of core components, assessing the real-world feasibility of each part to arrive at a surprising conclusion about the primary obstacle.
1
The Component Roadmap
Establishes the video's structure by stating that the complex concept can be broken down into a specific number of core components, giving the viewer a clear path to follow.
2
The 'Easy Win' Component
Details the first component, explaining its function and presenting it as largely solved with current technology to build initial momentum and viewer confidence.
3
The 'In-Progress' Component
Details the next component, showing it's more complex but that significant progress is already being made, escalating the sense of possibility.
4
The Central Hurdle
Introduces the most difficult component, framing it as the main technical challenge that requires a major breakthrough, which injects tension and raises the stakes.
5
The 'Essence' Components
Details the final, most advanced components that truly define the concept's unique character, moving from the practical to the philosophical.
6
The Unified Synthesis
Briefly summarizes how all the individual components work together in a single, cohesive system to create the final concept.
Themes
Technological advancement
Science fiction vs. reality
Future of AI
Techniques
Motion graphics
B-roll integration
Pop culture reference
Post caption
You know J.A.R.V.I.S from Ironman?
What if I told you, we’re actually way closer to building it than people realize.
To make J.A.R.V.I.S, all you need is an AI system that can do 5 things.
First, you have to be able to communicate with it.
It needs to hear you clearly and speak back fluidly.
But we already have AI tools that can do this really well.
Eleven Labs and Hume AI are world class voice models that are already at J.A.R.V.I.S’s level.
Second, J.A.R.V.I.S needs to be able to see things and visually understand what’s going on.
It can’t just have ears, it also, needs to have eyes.
And those eyes wouldn’t be through a single humanoid, it would be through a series of cameras all over your space.
But again, we already have world action models that are getting pretty good at seeing and understanding.
But the third piece is the biggest one…
In order for J.A.R.V.I.S to work, you need one central AI brain that can essentially hop around freely to different robotics and smart devices throughout your space.
It’s kinda like seamlessly switching sonos speakers from room to room, but on steroids.
Today our smart devices are actually pretty dumb.
So to have a legit working version of J.A.R.V.I.S, we’d need an interconnected system of devices and machines that is smart enough to activate as you maneuver around it.
So that’s the brain and body, still a ways to go.
Part 4 and 5 is really what makes J.A.R.V.I.S, J.A.R.V.I.S.
And that’s the customizable personality and the proactive action system.
On the personality front, we really just need better memory and emotional expression.
But this is also coming and the progress in memory is obvious in the latest versions of ChatGPT.
The proactive action system is having J.A.R.V.I.S assess and be able to do stuff without you asking.
This is essentially an AI agent system.
Ironically the biggest problem holding back J.A.R.V.I.S today is not that our AI isn’t smart enough…
It’s that our cameras and robots are too dumb….for now.
Follow Kane K. for more videos like this.
Tried something new with this one, lmk what you thought!
#ai #artificialintelligence #tech #technology #ironman #jarvis #tonystark #robot
Text overlays
title · beginning
J.A.R.V.I.S
contextual · beginning
from Iron Man?
informational · beginning
STEP 1
informational · beginning
COMMUNICATION
informational · middle
STEP 2
informational · middle
SEEING
informational · middle
STEP 3
informational · middle
INTEGRATION
informational · middle
Central A.I. Brain
informational · end
STEP 4 & 5
informational · end
CUSTOMIZABLE PERSONALITY
informational · end
PROACTIVE ACTION SYSTEM
informational · end
BETTER MEMORY
informational · end
EMOTIONAL EXPRESSION
informational · end
VISION
PROACTIVE INTEGRATION
PERSONALIZED COMMUNICATION
contextual · end
for now...
Transcript
You know Jarvis from Ironman? What if I told you we're actually way closer to building it than people realize? To make Jarvis, all you need is an AI system that can do five things. First, you have to be able to communicate with it. It needs to hear you clearly and speak back fluidly. But we already have AI tools that can do this really well. Eleven Labs and Hume AI are world class voice models that are already at Jarvis's level. So the listening and talking piece, that's kinda good to go. Second, Jarvis needs to be able to see things and visually understand what's going on. It can't just have ears, it also needs to have eyes. And those eyes wouldn't be through a single humanoid, it will be through a series of cameras all over your space. But again, we already have world action models that are getting pretty good at seeing and understanding. So Jarvis' eyesight capability is also a lot further along than people realize. But the third piece is the biggest one. In order for Jarvis to really work, you need one central AI brain that essentially can hop around freely to different robotics and smart devices throughout your space. It's kinda like seamlessly switching Sonos speakers from room to room better on steroids. Today, our smart devices are actually pretty dumb. So to have a legit working version of Jarvis, we would need rank interconnected system of smart devices that are smart enough to activate as you maneuver around them. So that's the brain and bad, still a ways to go. Part four and five are really what make Jarvis Jarvis, and that's the customizable personality and the proactive action system. On tier personality front, we really just need better memory and emotional expression. But this is also coming and the progress in memory is obvious in the latest versions of ChatGPT. Tier proactive action system is having Jarvis assess the situation and be able to take actions without you asking. And this is essentially an AI agent system. And again, companies are starting to lay the rails today for proactive agents. So what Jarvis really does is use eyesight to understand what's going on, on, proactively take action across an interconnected hardware network and then communicate what it's doing back to you in a personalized way. Ironically, the biggest problem holding Jarvis back today is not that our AI isn't smart enough, it's that our cameras and robots are too dumb for now.
