The updates keep coming. But at least you don’t have to shut everything down and wait for them to load, right?
Same, but different
Open AI just announced GPT-4, the multimodal version of its famous language model.
The latest iteration—which Microsoft admitted is already powering Bing chat—is “more reliable, creative, and able to handle much more nuanced instructions.”
Impressive
Besides being an overall better “verbal assistant,” GPT-4 brings plenty of new tools to the table, including…
- Understanding images. Take an open refrigerator, for example. Not only can GPT-4 identify what’s inside—it can also suggest what meals you might prepare based on the ingredients. Jeez.
- A new API system that can help developers “steer” the language model in different directions. In other words, setting the tone of the language model for a particular app.
Early adopters of the new model include DuoLingo, Stripe, and Morgan Stanley, among others.
You can test it out too, as long as you have a ChatGPT Plus subscription.
Two can play this game
Meanwhile, Google opened up PaLM—its language model—to developers. Likely to compete against OpenAI’s effort to help users build apps with GPT.
The PaLM API can carry out all sorts of text generation and editing tasks, allowing you to power your chatbot, summarize text, and even code. Yes, just like ChatGPT.
Not done yet
In addition to PaLM, Google has two more enterprise tools for users, namely…
- Vertex AI to help enterprise companies train and deploy machine learning models.
- MakerSuite app builder, a tool that allows you to quickly “ship new experiences” including bots, chat interfaces, custom search engines, and more.
Why we care
Each of these improvements could enhance your career, your performance, and inspire you to try new things, or even build them.
And we have to say, we’re excited to see what you’ll do.