Elon Musk Working On Humanoid Robots To Make Cars Safer
What’s In The Buzz?
As per the announcement made by Elon Musk, Tesla has started rolling out prototype humanoid robots by next year.
This confirmation was given at the Tesla AI Day event in California, where the company looked to recruit new machine learning talent.
- The event brought a new custom chipset whose task is to train artificial intelligence models at data centers.
- The new Dojo D1 chipset is going to be a part of Tesla’s supercomputer. It is equipped with 362 teraflops of processing power. Furthermore, it is built on 7nm fabrication.
Here are some more highlights
Tesla Bot- The Companie’s New Humanoid Robot
Elon Musk, the CEO of Tesla has stated the company to be one of the largest robotic’s companies. And the bot will leverage some hardware and software that will be used in the company’s autopilot driver assistance.
- He further claimed that the humanoid robots will be designed for handling unsafe, repetitive, and boring tasks, mainly helping the company in solving the problem of labour.
- He added that in the future physical work is going to be a choice.
- The website of the company also revealed that the company is looking for talent to bring the Tesla bot into reality.
- They are seeking help from mechanical, electrical, controls, and software engineers to leverage their AI expertise beyond vehicle feet.
Dojo D1 Chipset For Training AI Models
Tesla has put on the table its new chipset that has been named D1 for training the AI models that are used in the company’s data centres.
- The above is going to allow the Dojo computer for processing imaging data four times faster than other computer systems.
- The idea behind this is to push the AI models that are trained by the D1 chipset for existing Tesla cars through software updates.
- Another thing for building the chipset and having integration of hardware/software is to reduce the latency with improvement in bandwidth.
- They can perform both computing and data transfers, which is mainly the instruction set architecture that has been optimized for machine learning workloads and is pure machine learning.
Also read: Using a New AI Algorithm, MIT has designed Soft Robots That can Sense
Seld-Driving To Get Better This Year
The company is selling a full self-driving capability, which is an add-on. It lets the car of the company automatically change lanes while driving on highways, parking the car, or emerging from a parking spot.
- Further, the package will not come with the additional ability to automatically steer on streets.
- The company is also offering an add-on for a monthly subscription for those who cannot pay a lumpsum amount.
Expansion Of Self-Driving To Other Cars
In the process of elaborating the current status of full-self driving, Tesla hoped that the learnings from the supercomputer in the field of self-driving and can expand to other carmakers down the road.