Welcome back friends and new AI enthusiasts to ChildsPlayAI, a weekly newsletter where we break down what’s happening in AI, in language you can understand.
You’re favorite late-night and daytime shows are coming back. The writer’s strike is over and turns out, AI played a big part in why we were left hanging on recipes for our “perfect summer party.”
Also, how many times can you say CIA AI?
Who doesn’t like to wind down to some late-night shows, tune in during the day for some hot gossip, or get a mulled wine recipe from a previous American Idol contestant? At the very least, it’s some much needed entertainment.
If you’re like me, you’ve been missing some of this programming since the writer’s strike began in May. The good news is that as of this week, it looks like this strike is finally coming to an end. Aside from some of the more “normal” asks from the writers union (better pay, benefits, etc.), there is a major emphasis on protections for these workers regarding the use of AI. The key AI highlights of the deal are:
AI cannot write or rewrite literary material. This means that studios cannot use AI to generate scripts or rewrite existing scripts without the involvement of human writers.
Writers can choose to use AI if a studio approves of its use, but a writer cannot be required to do so. This gives writers control over whether or not they want to use AI in their work.
However, the agreement does not prohibit studios from training AI on writers' work. This means that studios can still use AI to generate ideas, learn about writing styles, and develop new tools to assist writers.
Overall, the agreement aims to strike a balance between protecting writers' jobs and creative control while also allowing studios to explore the potential benefits of AI.
While it is not clear how easily the agreement can be translated to other industries, there is reason to speculate that these terms could lead to larger implications and precedents. For example, unions in other industries, such as manufacturing and retail, could use the writer’s agreement as a template to negotiate their own protections against AI.
For now, I’m choosing to focus on the short-term win of having back some guilty programming pleasures. Rest assured though, the protections we’re provided in the workplace will become a hot topic as advancements in AI continue to automate the world around us.
The Central Intelligence Agency (CIA) has confirmed they are building their own version of ChatGPT AI to be utilized across the US intelligence community. This is being called a “logical technological step” in streamlining information across 18 different agencies ranging from the CIA, NSA, FBI, and several military offices, but as you can imagine, there are much deeper implications at play within this move. Let’s review some of their main use cases :
The same as a ton of other folks…to sift through massive amounts of data, but this is some powerful and sensitive data. Since 1947, the CIA has been collecting a vast amount of data from a variety of sources, including human intelligence, signals intelligence (monitoring communication signals from foreign agencies), and open-source intelligence (publicly available information). Now, with the help of AI, the CIA will be able to analyze this data more quickly and efficiently and identify patterns and connections that would be difficult or even impossible to find manually.
To develop new intelligence gathering and analysis techniques. AI can be used to develop new ways to collect and analyze intelligence, such as by using natural language processing to extract insights from social media posts, or by using machine learning to identify patterns in satellite imagery. For instance, this could be used for tracking terrorist networks and to detect cyberattacks.
To stay ahead of adversaries and maintain its intelligence advantage. China and other countries are investing heavily in AI for intelligence purposes, and the CIA’s AI Director is calling the new tech our biggest threat.
As you can imagine, the general public perception of the CIA building its own AI is mixed. There are concerns about the potential for misuse, but there is also recognition that AI is a necessary tool for the CIA to maintain its intelligence advantage and to protect national security. Regardless, this will be a story we should all want to keep an eye on moving forward.
Chat with a PDF. No joke.
We’re totally in the future now. Because you can chat with a…document. That’s right folks. PDF.ai allows you to upload any PDF and have a conversation with the document. Ask any question you like. PDF.ai brings your documents to life. Applications here are really endless - books, legal documents, employee handbooks, user manuals. Now go impress your new boss about how much you know from the employee handbook on day 1.
We could use a little hand here…..