Keep pulling the thread on Eric Landau.
Some experts in the robotics field predict there will be more robots than people within 5 to 10 years.
Encord's competitive differentiation is its focus on physical AI and its ability to operate on multiple petabyte-level datasets.
The "bitter lesson" in AI development is that scaling a system with more data and compute is more effective than feature engineering and inducing domain expertise into the system.
The release of ChatGPT helped Encord by changing the tenor of conversations around AI and making the market more open to their product.
Encord's co-founders realized they had achieved product-market fit when a sale closed in their Gong channel from a company whose business they were not familiar with.
Encord went through two unsuccessful sales teams before their third attempt at building a sales organization started to work.
Encord's initial product strategy was to start with computer vision, believing it was the hardest data modality, with the long-term goal of becoming a multimodal platform.
The biggest application of multimodal AI is now in physical AI, which includes robotics, autonomous vehicles, logistics, and manufacturing.
An estimated 80% of economic activity involves manipulating or moving things in the real world.
The biggest growth area for Encord is currently in robotics, including humanoids, consumer robotics, and manufacturing.
Encord operates a facility in the Bay Area with robot systems and human operators to collect data for various tasks.
Encord's product offering includes data management, curation, annotation, enrichment, and evaluation tools.