FTC Probes OpenAI

Welcome to another edition of the best damn newsletter in AI. Here we’ll break down AI topics that matter, open your mind to use cases, and keep you ahead of the curve.

Our #1 goal is to be useful. So please shoot us an email 📩 if you have questions or feedback, and especially if you implement something we share!

Here's what we're covering today:

• Takeaways from the FTC move to investigate OpenAI

• Using Code Interpreter to clean data

• Hollywood strikes while news publishers lean in - an update on content and data training policies in AI

Let’s get to it! 👇

TODAY'S PERSPECTIVE

The FTC Opened An Investigation into OpenAI

What does that mean? And is this a wake-up call for people using AI? Here's what you need to know.

The Federal Trade Commission (FTC), which is in charge of consumer protection, began an investigation on OpenAI. It appears to be at the civil/information gathering phase, but could certainly progress to something more formal and serious.

But when we read the investigation letter - we were quite surprised by its depth already.

The FTC is pressing OpenAI on many issues that are top of mind: what data were these models trained on, who has oversight and responsibility, how user data is used, hallucinations, and more.

While the investigation is unlikely to have near-time implications for any of us, we think this is important to watch to understand AI adoption, how to use AI in your work, and especially if you're building AI products or services for customers.

Here are the key takeaways: 👇️ 

1) AI is now a consumer protection issue

There have been enough reports on hallucinations, data leaks, training data concerns and more that have regulators worried.

Our team has worked in highly regulated environments in the past - some regulation is often a good thing. Although it will slow down innovation and adoption, which carries its own set of risks in the global rush to adopt AI.

2) As AI adoption becomes more widespread, so does the responsibility of ensuring data security.

The combination of the fast pace of AI adoption, how young most of the companies in the space are, and the sensitive nature of our own data being used to train these models has certainly put people on edge.

And if you’re building a product or service using AI, you need to be aware and cognizant of that.

3) Even though the government is stepping in on data privacy, it might be a while before we have solid regulations

So again, create your own data policies for yourself, your work and your businesses.

And if you haven’t already, get our free guide to data privacy and ownership.

4) AI adoption is happening so fast that compliance can’t really be an afterthought

While this post isn’t sponsored, we do want to highlight that if you’re serious about building AI products or services for people - you should be paying attention to data governance and compliance regulations.

We personally like Secureframe as a leader in content and tooling in the tech compliance space.

Premium Members: We also have a workshop this week with Secureframe this Thursday, July 20, from 1-2pm PT: Grab your Spot. 

It’ll be a great event for anyone interested in learning more about compliance certifications like SOC 2, GDPR, etc. Nerdy but necessary 😉 

USE CASE

Code Interpreter Use Case #2: Tell your data to clean itself

Due to popular demand from our last code interpreter use case - we’re going to walk through a series of use cases over the next few weeks.

Today’s use case: cleaning data 🧹 

Rachel’s a former data scientist and she’ll tell you — the worst part of the job is having to clean your data. And seriously - with Code Interpreter you can now just say what you want.

For example:

1. Upload your raw data to Code Interpreter

2. Ask the model - “what are some data cleaning tasks I should do before starting my analysis”

3. Ask it to do some of the tasks that it suggested

4. Download your clean data once done!

5. Tip: We recommend using one Code Interpreter thread to clean your data, and then using a different one for specific tasks to save on memory 🙌 

Here’s a peek at some of the data cleaning tasks it suggested for us in a recent task 👇️ 

and that’s it!

Premium Members: Read our full playbook chapter on Code Interpreter here + our tutorial on how to use Code Interpreter to analyze public financial reports to make better business decisions 🔥 

LINKS

For your reading list 📚

AI features keep coming and the functionality keeps getting better and better…

• Google’s Bard now can accept images, meaning you can talk and respond to visual prompts

• Wix is entering the game - they announced they’ll be adding an AI-powered website builder to the various ways they offer to create custom websites

On the content licensing front…

• OpenAI and the Associated Press announced a partnership where they will use the AP’s articles to train their models

• Meanwhile Hollywood actors are still on strike at the proposed usage of AI replicas based on actors’ likeness

And if you're really nerdy...  

• The SEC warns that relying on AI for financial markets could cuase market instability, and encourages people to visit their Fin Hub for information on finance, crypto and AI

That's all!

We'll see you again on Thursday. Thoughts, feedback and questions are much appreciated - respond here or shoot us a note at [email protected].

Cheers,

🪄 The AI Exchange Team