AI is going to be changing the workforce and we want to surface the best opportunity for early adoption. We have been reviewing AI marketplaces, infrastructure tools for deep learning developers, and tools for non-developers to leverage the benefits of AI. These models are heavily reliant on access to data and numerous developers building, training, and in some cases sharing machines. More open data sets create more opportunities to start machine learning training.
Data comes from a few places: developers, existing data sets, or scraping/cleaning data from other sources. Developers train on these platforms, build and share so others can replicate/use, and create additional data feedback into the ecosystem. The more people that run the platform’s models or train their models on the platform's data, the smarter the machines get for both the platform and the developer. Marketplaces and model builders require more feedback loops of data into their ecosystems to strengthen them. Building the right concentration of developers and sharing is important to allow more innovation in this field.
We continue to evaluate companies in this space, weighing the team, focus, tech, and target users. Companies are either too early in their concept but we are keeping an open door and others are not differentiated enough in their thinking from existing competitors. We believe the team that is focused on providing accessible machines to the right market will be the winner.