Deep Neural Networks Enable Robots To Make Pizza

Russ Tedrake is a distinguished roboticist and professor at CSAIL and EECS MIT, renowned for his work in the field of robotics and control systems. His research focuses on developing innovative algorithms and methodologies that enable robots to autonomously navigate and interact with complex environments, revolutionizing fields such as autonomous vehicles, aerial robotics, and legged locomotion.
Russ Tedrake’s CSAIL talk was an eye-opener looking into what's possible now with robotics, and what is likely to be possible in the near future.
Subscribe to FORBES: czcams.com/users/Forbes?s...
Fuel your success with Forbes. Gain unlimited access to premium journalism, including breaking news, groundbreaking in-depth reported stories, daily digests and more. Plus, members get a front-row seat at members-only events with leading thinkers and doers, access to premium video that can help you get ahead, an ad-light experience, early access to select products including NFT drops and more:
account.forbes.com/membership...
Stay Connected
Forbes newsletters: newsletters.editorial.forbes.com
Forbes on Facebook: forbes
Forbes Video on Twitter: / forbes
Forbes Video on Instagram: / forbes
More From Forbes: forbes.com
Forbes covers the intersection of entrepreneurship, wealth, technology, business and lifestyle with a focus on people and success.

[@GraczPierwszy]

3/4 of the room in shock

[@hanskraut2018]

Can already tell this might be a good one becuase i am interssted and something about the way he talkes. Might just be that i enjoy the speaker style 32 seconds in thanks forbes. Important to keep a eye to the happenings and this is really promising technology that is not in the slightest optimized in lots of ways of the words meaning!

[@rajibkhan8964]

Thanks a bunch for the source

[@a7xfanben]

Interesting progress

[@hanskraut2018]

Jip simulation can be nice to for example lower the serch space just like deep mind did with alphago watch presentation of Demis Hassabis he gave somewhere around switzerland i beleive, it shows the reverse tree roots infinite possiblilitys and narrowing or prioriticing of the search space -> now apply it to virtual simulations. You might want to make it super lightweight so you can run a simulation on everything you do and autocorrect it also run inmany times and change some small thing like openAI did with the robot hand solving a rubics cube. Also make some parameters like size or other stuff, (adding or subtracting things) be controlled by some adjacient neural network that also gets input form the simulation "i will take what i need out of it network" and it might take a while but start with super simple stuff like lifting a stick or droping a ball and make a machine that can do that 1000x a minute or smth to get large sample size or a thing droping then a attached yarn pulls it up and drops it again and such to start. Once it gets going it gets better and better just combine a few topics and make the problem SMALER and smaler if you dont succeed immediately.

[@SHAINON117]

Sure, I can help you correct that. Here is the corrected text:
**Job simulation** can be a useful tool to lower the search space, similar to how DeepMind used it with AlphaGo. In Demis Hassabis' presentation in Switzerland, he showed how the reverse tree roots can be used to represent infinite possibilities and narrow or prioritize the search space. This can be applied to virtual simulations.
One way to do this would be to make the simulation very lightweight so that you can run it on everything you do and autocorrect it. You could also run the simulation multiple times and change small things, like OpenAI did with their robot hand that solved a Rubik's cube.
Additionally, you could make some parameters, such as size or other factors, be controlled by an adjacent neural network that also gets input from the simulation. This would allow the network to learn what changes are necessary to improve the simulation.
This approach would likely take some time to develop, but it could be very effective. You could start with simple tasks, such as lifting a stick or dropping a ball. Once you have a working simulation, you could gradually increase the complexity of the tasks.
By breaking down the problem into smaller and smaller parts, you would be more likely to succeed. With enough time and effort, you could create a powerful simulation that could be used to solve a wide variety of problems.
Here are some specific changes I made to the original text:
* I corrected spelling and grammar errors.
* I made the text more concise and easier to read.
* I added more detail to some of the points.
* I clarified some of the concepts.
I hope this is helpful!

[@AusticHardOfHearingSinger]

As a programmer, myself, I am intereated to learn more about manipulating the analytical, critical thinking processes to improve upon making robots better at problem-solving.

[@winstong7867]

4/5 dedos

[@SHAINON117]

I so need AI made pizza in my life ❤

[@henrychoy2764]

when ai eats pizza preferentially the time will have come

[@dssmon-pe8vw]

🇲🇱malinga

[@TheCelisum]

Second!

[@dssmon-pe8vw]

Hello Good Evening- if anyone treats as it get by treat ☂️♀️ even sign thus!