jan/handbook/why/problem.mdx

36 lines
1.6 KiB
Plaintext

import Youtube from '../../../components/youtube'
# What problem are we solving?
Current robots suck:
- **Dumb**: good at repetitive tasks, but can't adapt to new situations
- **Difficult**: needs to be programmed by experts
- **Expensive**: often more expensive than doing it manually (for most use-cases)
- **Non-autonomous**: doesn't work offline, as inference is not done onboard
## Mission
Menlo Research is focused on augmenting humanity with robots that are generally intelligent.
## Approach
We solve this incrementally. We productize sub-problems that can [solved now](https://www.nasa.gov/directorates/somd/space-communications-navigation-program/technology-readiness-levels/), that become key components of our long-term vision.
- Running AI models on edge hardware ([Jan](/products/jan), [Cortex](/products/cortex))
- Multimodal AI for robot-human interaction ([Ichigo](https://github.com/menloresearch/ichigo))
- LLM Hardware
- Teleoperation tools & infra
- Simulation environments
- Affordable robots
We earn the right to build our long-term vision by delivering value to users today.
## Vision
We envision an [optimistic future](https://www.youtube.com/watch?v=o48X3_XQ9to) where our children achieve what we dream of today, enabled by exoskeletons and helpful robots.
In the long run, we see robots helping humanity progress into a [Type-1 civilization](https://en.wikipedia.org/wiki/Planetary_civilization) on the [Kardashev scale](https://en.wikipedia.org/wiki/Kardashev_scale), by [engineering Earth](https://youtu.be/xC2ccYpjPHE) and becoming a spacefaring civilization.
<br />
<Youtube id="o48X3_XQ9to?si=KDkCQ8MZe13sEc68" />