Welcome to another edition of the best damn newsletter in AI.

As a reminder, we love Reader Questions. Our mission is to make using AI accessible to every business, and a huge part of that is making sure you’re not stuck and that we continue answering diverse questions. So, submit your question here. We’ll review every single one and if we believe i’ll be helpful, we’ll cover answers here in the newsletter.

Here's what we're covering today:

• 3 More Ways on How to Plan for New OpenAI Models

• Become a better writer with AI feedback

• Technology designed to help AI understand human emotions

Let’s get to it! 👇

TODAY’S PERSPECTIVE

3 More Ways on How to Plan for New OpenAI Models

OpenAI is rumored to be releasing their newest AI model (nickname Strawberry) in the next 2 weeks.

The general consensus is that this model is supposed to be better at reasoning and math than its predecessor GPT4.

The rumors state that the way it does this is by "going away for 10-20 seconds, thinking, and coming back with a more robust answer".

While this may sound exciting for anyone frustrated with AI making things up, especially math (!!), it's very unlikely this will make ChatGPT reliable enough at math on its own.

Here’s why and how we are thinking about it instead:

They are still Language Models, not Math Models

ChatGPT and the upcoming “Strawberry” are still language models, fundamentally designed for language, not math. 

They understand numbers as language when processing them, which is akin to guessing answers based on how numbers sound rather than precise calculations.

This inherent design flaw means they can still produce inaccurate results when performing complex math tasks.

Using ChatGPT to create data analysis code is more reliable

This doesn’t mean don’t use ChatGPT for math. In fact, one of the best use cases is to get ChatGPT to generate data analysis code that does the math – rather than doing the math itself.

Instead of relying on AI for direct math calculations, we use it to write Python code that runs the data analysis (or write excel formulas, or macros… the same concept applies).

Having AI write code to do data analysis, rather than doing the analysis directly gives you more control and can ensure accuracy and consistency each time you run it. 

If you want to learn more, we like this tutorial: How to analyse data using ChatGPT and pandas

Pro tip: consider using ChatGPT to explain the code to you, check its own work or make changes to make the data analysis even more bullet proof!

Plan on using the newest models for more advanced reasoning

In our work, we are constantly asking ChatGPT to think through complex situations, briefs, ideas or analysis.

We expect that models like Strawberry will make it even easier to teach AI to reason and think through problems like we do. That means its even more important to get specific and articulate what you want the AI to reason through or HOW you want it to complete the task. If you do, chances are, it will be able to follow it even better with the newest models.

And that, we are very excited about.

USE CASES

Become a Better Writer with AI

AI tools like ChatGPT are great for getting your ideas down quickly, but let's face it—they often produce average results that need some polishing.

And if you’re not a writing pro, it can be tough to know exactly what to tweak to make your content great.

If you’re just getting started with using AI for writing feedback, OpenAI’s Professional Writing Coach GPT is a great place to start.

This GPT basically acts as your personal writing coach. It reviews your text and provides constructive feedback to help you refine your style, clarity, and tone. Here’s how it works:

1. Head to Professional Writing Coach GPT on the ChatGPT platform.

2. Paste in your email, article, or memo—whatever you’re working on.

3. Ask for specific feedback on grammar, style, or overall effectiveness.

4. Instantly receive tailored suggestions to elevate your writing.

Check it out and let us know what you think!

LINKS

For your reading list 📚

Funding and Investments in AI

• Amazon has started testing ads in its shopping-focused chatbot, Rufus.

• OpenAI co-founder Ilya Sutskever's new AI startup, Safe Superintelligence (SSI), raises $1 billion in funding to develop safe AI systems that surpass human intelligence.

• You.com, an AI-driven search engine, secures $50 million in funding to enhance its ability to answer complex questions.

AI in Consumer and Enterprise Products

• Zillow upgrades its AI search feature to allow users to find homes based on proximity to specific locations.

• Volkswagen launches an upgraded in-car AI voice assistant, IDA, integrating ChatGPT and other cloud-based AI models.

• Canva increases the prices of its Canva Teams subscriptions due to the addition of new generative AI features.

• Apple is launching a smarter version of Siri this fall, enhanced by its new AI initiative, Apple Intelligence, and a partnership with OpenAI.

• NightCafe, an AI-powered art platform, has built a strong user base of over 25 million people, offering a range of image-generation tools through an easy-to-use interface.

AI Ethics, Art, and Human Interaction

• Science fiction writer Ted Chiang argues that generative AI will never create true art, as it lacks the human ability to make meaningful choices and engage in genuine communication.

• "Emotion AI," technology designed to help AI understand human emotions, is gaining traction as businesses integrate AI into various roles.

Have a burning AI question? Or stuck trying to help your team adopt AI?

We read every single question submitted and use these to inform the topics we go deep on. 👇️ Submit your question below!

That's all!

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

Cheers,

🪄 The AI Exchange Team