Last year, San Francisco-based AI lab OpenAI released GPT-3, its latest attempt at a program that writes like humans.
To do this, the program is trained using 175B parameters. For context, all of the English Wikipedia is estimated to make up just 0.6% of the training data.
Now, OpenAI is looking to bring in the dough
The lab was founded as a nonprofit in 2015, but created a for-profit offshoot in 2019 to help drive funding, including a $1B investment from Microsoft.
Now, just 9 months after GPT-3’s launch, 10s of thousands of developers are building on the platform and the API is generating ~4.5B words per day.
Already, >300 apps are using GPT-3 for everything from productivity apps to games:
- Viable is using it to help companies summarize customer feedback
- Fable Studio is creating GPT-3-powered interactive video game characters
- Algolia offers customers GPT-3-driven semantic search capabilities
But there are hurdles ahead
GPT-3 itself isn’t always the sharpest tool in the shed. In one instance, a GPT-3-powered medical chatbot encouraged suicide.
Some worry that OpenAI’s leaders quietly envision a world in which they run every non-AI company out of business. On the surface, that doesn’t entirely align with the group’s mission to benefit “all of humanity.”
However, the group’s CEO has said AI can soon generate enough wealth to pay every adult in the US $13.5k annually. That doesn’t sound too bad.