Microsoft has announced that it will offer access to AI models that form part of ChatGPT and OpenAI to U.S. Government Agencies using Azure Government. Users will be able to take advantage of both GPT-3 and GPT-4. Let’s take a moment to catch our breath! What is ChatGPT? Who’s using it? Is it any good? Should we be worried about the US government using it?
What is ChatGPT and who’s using it?
Microsoft has stakes in ChatGPT, who have led the charge in the recent surge of large language models. Competitors, like Google’s Bard, along with Nvidia, DeepMind and Meta are fighting to dethrone OpenAI’s ChatGPT. Since shooting to stardom early this year, it’s been used by Cambridge students to write their essays, a lawyer submitted a brief not realising it was filled with fake citations and Charlie Brooker recently got it to write a Black Mirror episode. He hated it!
There’s a lot of excitement about this new technology, it appears to be able to do anything but its limitations and dangers are becoming more apparent. A former OpenAI researcher even warned recently that it has the potential to “annihilate” humanity. The threat to screenwriting at least, has formed part of the argument for the ongoing WGA writer’ strike.
Does It Live Up To The Hype?
Even if its results aren’t so catastrophic, it’s been pointed out that there are certain things it can’t do. A group of scientists have developed a test that can determine with total accuracy whether a science paper was or wasn’t written using ChatGPT. They point out that human writers use a more diverse vocabulary and creative style of writing than the chatbot. Their strongest critique however, is its inaccuracy.
Heather Desaire, a chemistry professor at the University of Kansas writes…
“Right now, there are some pretty glaring problems with AI writing”…”One of the biggest problems is that it assembles text from many sources and there isn’t any kind of accuracy check – it’s kind of like the game Two Truths and a Lie.”
While these things may sound alarming, we shouldn’t write it off. ChatGPT’s incredible rise in popularity, doesn’t come from nowhere. Maybe we shouldn’t entirely rely on it for our writing but it can help with a whole lot of tasks, like providing frameworks for how to tackle a piece of writing or research, or untangling a mess of information. It’s proven particularly helpful for small businesses.
Knowing all this, how do we feel about this tool being used by U.S Government Agencies? This will be the first time such a thing has been made available to government. It’s thought that the technology will be used at the Defence Department, the Energy Department and NASA. The Defense Technical Information Center, where researchers gather and share military information also plans on utilising the AI models. What do you think? Are we ready to wholeheartedly embrace ChatGPT? Let us know in the comments!