One of the kerfuffles of the week has been Bing’s unpredictability. This paragraph from the Verge sums it up. In conversations with the chatbot shared on Reddit and Twitter, Bing can be seen insulting users, lying to them, sulking, gaslighting and emotionally manipulating people, questioning its own existence, describing someone who found a way to force the bot to disclose its hidden rules as its “enemy,” and claiming it spied on Microsoft’s own developers through the webcams on their laptops. And, what’s more, plenty of people are enjoying watching Bing go wild.
Microsoft themselves responded in a blog post, observing that long chat sessions can cause issues, and it didn’t “fully envision” people using its chat interface for “social entertainment” or as a tool for more “general discovery of the world.” Extended chat sessions with 15 or more questions can confuse the Bing model.
Testers have already reverse-engineered the prompts and rules driving the Bing AI – identifying its name as Sydney. Microsoft confirmed the secret rules, despite one of the rules being that they are confidential and permanent.
With all this, the company also updated on the progress – testing with users in 169 countries and with a waitlist now in the multiple millions.
And I’ve already spotted the first “Prompt Engineer” role. Paying between 250 and 335 thousand dollars, the role is to “figure out the best methods of prompting our AI to accomplish a wide range of tasks, then document these methods to build up a library of tools and a set of tutorials that allows others to learn prompt engineering or simply find prompts that would be ideal for them.”
NICE InContact has announced that their CXone Expert now uses OpenAI’s generative modeling. Otter.ai has also announced its new AI-powered meeting summary tool, OtterPilot.
And in one more for the list of things AI can do, the DOD’s research agency has announced that it’s AI Algorithms can now control an actual F-16 fighter aircraft in flight, flying multiple flights over several days. The goal isn’t to replace the pilot – instead, to have the “human pilot focuses on larger battle management tasks in the cockpit” and have the AI control the jet and provide live-flight data.
That’s what’s key – the combination to create more versus replacement. Axios covered how this is trickier than it sounds and cited specifically autonomous and semi-autonomous vehicles. Insider looked at writing and noted that ChatGPT couldn’t replicate creativity or empathy, nor could it beat spam policies.
Why do we care?
I have two insights to offer today. First, prompt engineering. Learning to master the prompts provided to an AI to get results will be an entire space to master and teach. This is a skill of the future, and I wouldn’t assume this one will just be easy to acquire, let alone train. Just don’t ask for five years of experience on the resume for this one. I’m putting my highlighter around this entire skill set as the consulting of the future.
The other was the idea of augmentation. Experts keep highlighting how AI will augment workers rather than replace them. I’m on board with this directionally, highlighting how difficult that is. I like complex problems due to their profitable nature. Again with that highlighter to draw attention to the need.