The AI API : The Twilio Moment for Machine Learning

90% of startups have plans or have released an AI feature, 54% of those features will launch in 2023, but only 30% of companies are hiring new people to do it,?according to Productboard’s survey.

These figures highlight three points:

1. AI has become?an essential product component for most software companies.
2. Startups have aggressively prioritized these features on their roadmaps with 54% launching a feature this year.
3. AI deployment is sufficiently straightforward that a majority of teams won’t hire new experts to build them & will staff 1-2 people to launch them.

The last one is the most striking.

Training, deploying, & optimizing machine learning models has historically required teams of dedicated researchers, production engineers, data collection & labeling teams. Even fully staffed, teams required years to develop models with reasonably accurate performance. Today, it’s a matter of days or weeks.

Newer LLMs (large-language models) simplify AI-feature development in the same way Twilio empowered developers to send an SMS with five-lines of code, abstracting the complexities of carrier-integration. Stripe did the same, simplifying credit card payments.

Building a chatbot for some collection of customer support articles or blog posts is straightforward :

1. Gather your documents
2. Save them in a vector database
3. Upload the vector database
4. Start chatting away by hitting the API

The time to utility is measured in minutes.

Every?company will be an AI company, because the technology is simple-enough to deploy & powerful-enough to work sufficiently well out-of-the-box that most companies can get started today.