Building A Remote Control T-Rex Robot
--
The purpose of this article is to demonstrate what I believe a better approach to creating a business plan will be in the future given the potential for AI to vastly accelerate entrepreneurs.
My goal in writing this article is to show how I believe “being able to write a sci-fi story” is going to be the single most leveraged activity for entreprenuers given the rapid progression of AI.
In the future, I believe that entreprenuers will be able to write stories like this, detailing key elements of their plan, then simply feed them into AI iteratively to eventually achieve the end result.
The only thing that matters is the willingness to outline your vision in explicit detail, then repeatedly iterate the vision with AI as AI progresses. So here we go.
WHY A TREX ROBOT
I have long wondered why it is not possible to purchase a believable, life-lIke bipedal dinosaur toy. In researching this topic, I have found that there are significant issues around balancing the robot and giving it a lifelike animation. 2–3 years ago, the best you could do was either a very clunky plastic toy robot or something along the lines of this Ostrish Robot.
It wasn’t until Disney introduced their biped in the last few months that I realized a bipedal dinosaur is now possible.
So why a T-Rex? Because it would be awesome, and it is now achievable.
DESIGNING THE TREX
The theory I have expressed in a previous article is that the future now belongs to “Story Engineers.” Basically: If you can express a convincing story and think through all the details, you will be able to achieve that story within the next 12–18 months by simply feeding it into successfully more capable versions of AI which will be built in this time period.
Whatever you can imagine, detail and write — you have a good chance of achieving it. So why not start the process with a T-Rex toy? This article IS the entire product, business, supply chain and engineering effort to achieve my vision of a convincing T-Rex toy. So let’s do it.
MOCKING THE MECHANICALS
Current ChatGPT and DALL-E are not capable of designing working mechanicals. But that is going to change in the next 3 years or so. So to begin, I walked through the process of asking DALL-E to design various components such as:
- Legs of the robot
- Batteries
- CPU
- Gyroscopic stabilization
- Hips
MOCKING FUTURE PROMPTS
So here is the thing. We know DALL-E is not (today) capable of converting prompts into working mechanicals, but we are going to assume that in 18 months, it will. So lets start with prompts which we think we will need and just assume that AI will figure it out, eventually.
You get the idea. None of these are remotely going to be something anyone can produce. However, in another 18 months, my expectation is:
- I can ask the AI to go source batteries, components, gyros, wires from Amazon or the internet
- Use these components, along with CAD / 3D printing to design hips, legs, joints, head, tail
- Produce 3D-printable materials so I can construct the robot
- Generate an instruction book about how to put them together
The purpose of going through this exercise is to simple “beta test” the ultimate design which will eventually be possible.
THE BUSINESS PLAN & GTM
So what would people do with 3D printed, home made T-Rex toys which are generated dynamically via AI using off-the-shelf components? I asked the AI to create some images. I imagine hobbysts would want to meet in garages and race their bipedal dinosaurs.
I imagine that creating the T-Rex robots would be a fun hobby for children to do with their parents to learn about the process of manufacturing on demand robots with AI.
BATTLE BOTS USE CASE
Maybe it would be fun to build robot T-Rexs and make them fight. I can imagine once people master the basics of building the home T-REX robots, they could scale things up and make them fight kind of like Battle Bots.
GETTING THE TREX TO WALK
A tricky part of the process would be to train the T-Rex to walk, I imagine that having a treadmill with a harness would be necessary to validate the concept.
I am also imagining using gyroscopic stabilizers to prevent the T-Rex from tipping over left and right.
THE TREX LAB
To make all this come together, more likely than not, you would need to create a warehouse complete with treadmills, 3D printers, race track and components to test out the bipedal robots.
BATTLE BOTS WALKING GYM
You would need a sort of “Robot Gym” to facilitate testing the robots and walking them through all the potential scenarios. The larger T-REX used for the Battle Bots scenario would require custom treadmills and harnesses.
GYROSCOPES
Getting the TREX toys to pivot and navigate sharp corners would involve the development of a custom gyroscopic stabilizer component which spins at high speeds and prevents the TREX from falling over.
Engineers would need to find clever ways to leverage the effect to prevent the bipeds from tipping over.
FUNDING, ACQUISITION & DISNEY DEAL
The most likely avenue to generating a business might involve licensing deals with Jurassic Park franchise, toy lines, perhaps deals with movies like Toy Story as well.
OPEN SOURCE TREX COMMUNITY
Once the initial open source mechanicals, robot software, bill of materials are designed and published, along with a simulator to validate designs, the open source community would print out their designs at home and build them as a community.
SIMULATOR
The TREX walking simulator would be produced by ChatGPT using Unity and be useful for pre-validating designs before the process of production.
TREX PERSONALITY STORE VIA GPT
Members of the open source community would design and train a range of GPT-based TREX personalities which can be downloaded into the TREX brain unit (probably off the shelf Raspberry Pi or similar).
TREX can have different personalities and learn new behaviors.
SOCIAL MEDIA
Recording their pet TREX doing different crazy things on TikTok and YouTube would be common and popular.
CLOSING THOUGHTS
Once again, the purpose of this story is to model the future. I believe AI will progress rapidly to the point where 100% of what I have documented here is not only possible but actively in-market in 18 months.
In other words, the value of being able to simply outline an idea like this end-to-end with the expectation that AI will enable it is the value here.
My expectation is that I can feed this document into AI in 6, 12 and 18 months and produce nearly everything I have imagined here including the simulator, shopping list, 3D printing designs, and more.
