It’s been almost a year since OpenAI introduced GPT-3, the language prediction model. TechCrunch called it “the tipping point for artificial intelligence” noting that it will soon power thousands of new startups and applications. That's probably not an exaggeration.

As a quick refresher, GPT-3, or Generative Pre-Trained Transformer 3, is trained on 175 billion parameters making it the largest transformer-based model in the world at the time of debut (Google researchers have since spoken about their Switch Transformer model that they say has been trained on 1.6-trillion parameters). As a transformer, GPT-3 is a deep learning model that mimics cognitive attention to decide what’s more important at every step as it processes input sequences.

Since its introduction, we’ve seen a lot of products launch that leverage all this artificial intelligence to develop new ways to process language-based tasks faster. A few have already gained great usage and fundraising traction (CopyAI just raised $2.9 million), and even though we’ve covered this before, there’s as much innovation as there are requests to hear what’s new in GPT-3-powered products.

Voices: Use AI voices to run your help desk or create ads

A/B testing: Optimise faster with A/B testing on autopilot

Sales emails: Get help with your sales emails and follow-ups

Fundraising emails: Or get help with pitching to investors

All the emails: Turn bullet points into an email in your own style

Long-form: Use Jarvis or ContentBot to write blogs and guides

Hiring: Create job descriptions and resumes faster (is the future of hiring all AI)?

Growth: Generate personalized growth strategies for your startup

Chatting: These shape bots have personality; Lila is a deep thinker

For those of you who love an open-source project, GPT-Neo is a grassroots project to create an open-source GPT-3 alternative.