Hey Sam,
Congratulations on finally launching your ChatGPT store!
At Groq® (with a q, not a k) we’re an AI technology company too. We understand how sometimes ship dates can slip. So what if you didn’t open the store in time for us to knock out our holiday shopping there? Everyone got socks instead. At least a January opening is just in time to help us pursue our New Year’s resolution of reading more books.
But slipping is one thing, slow is another. And unfortunately, using the custom GPTs that populate your new store is about as slow as trying to read War and Peace late at night.
We need some good books, so we asked one of the custom GPTs in the store for a suggestion of what to read. After proposing The Shadow of the Wind by Carlos Ruiz Zafón, the GPT suggested we consume the book while enjoying a cup of tea or coffee. We’re way ahead of you, since we had time to brew up a latte as your bot generated … its … response … one … word … at … a … time.
Let’s take that book somewhere sunny and warm, we decided. So we asked another custom GPT for recommendations of places to go that we could get to quickly. Its first response was the lovely town of Ojai, a bit over 300 miles south of here. We jumped in the car, zoomed out of the Bay Area, and pulled into our Ojai hotel just as the bot was completing its final suggestion.
OK, slight exaggeration (the bot was done somewhere around King City), but we hope you get the point. Speed matters. It drives interactivity and engagement. You can have the best model in the world, but if it’s too slow people will lose interest and move on.
We’re kind of speed freaks at Groq. Our LPU™ Inference Engine runs LLMs and other generative AI models about 10X faster than any other AI inference solution, which means that everyone could have real-time, engaging, back-and-forth conversations with the terrific custom GPT bots that populate your store. Which is, we believe, what AI can be all about: helping people engage, learn, and create, all at human speed.