869 Words | 3 min 40 Sec Read

Welcome to another issue of Geek AI.

Today we'll be discussing why Anthropic's new Claude 3 AI is being labeled as the most intelligent AI yet, beating both ChatGPT 4 and the newly released Gemini 1.5 on a variety of tests.

More important, we'll discuss what sets this under-the-radar apart from its competition and whether you should consider using it.

Let’s dive in.

This isue is sponsored by us - Geek AI

Overwhelmed by all this AI stuff? No idea where to start?

Meet AI Beginner Bootcamp: A once/week live training program where we'll show you how to use today's most important artificial intelligence apps.

From ChatGPT to Midjourney, our coaches will take you from beginner to advanced using simple, clear language. No tech jargon allowed.

Interested? Click here to learn more!​

Earlier today Anthropic announced their newest model: Claude 3.

While Claude never gained the same level of popularity as ChatGPT or Gemini/Bard, heavier users are familiar with it via their integration with AI aggregator Poe.com.

What was so impressive about today's announcement was that Anthropic labeled it the most intelligent AI on the market. But unlike most marketing plays, they had hard data to back up their claims.

In particular, as it relates to standardized tests used to gauge the accuracy, intelligence, performance, etc. of different AI models.

For the average consumer - who doesn't understand how these tests are run - it's near impossible to interpret what the results mean (beyond the fact higher scores = better).

What most people can understand, however, are the unique features and capabilities that make Claude superior to its competitors.

Sadly, Anthropic seems to have fine-tuned Claude 3 for Enterprise clients/business use cases. And because of that, it's questionable whether casual AI users will benefit from the upgrades.

This reality is further reinforced by the fact that - unlike OpenAI, Google and Microsoft - Anthropic is not selling a flat fee monthly package.

Instead, they'll be charging on a per-token basis (which implies they expect customers to access Claude via API instead of a website/app interface).

Either way, based on their release announcement, this is one powerful AI.

Here's just a small sample of what makes Claude 3 so powerful.

#1 - Conducting Complex Financial Forecasts

While everyday consumers use AI for light-hearted use cases like text and image generation, businesses have much more demanding needs.

Especially as it relates to processing large volumes of data. Which, as of now, ChatGPT is not very good at.*

*Google's public-facing Gemini app does not allow users to upload CSV files, but that functionality may exist for private small business/Enterprise clients.

Either way, we've had multiple ABB members express interest in finding an AI that can help them run complex, numbers-based data analysis.

And for now, it looks like Claude 3 may be the best candidate.

#2 - Task Automation

Ever since autonomous GPTs went viral last year, there's been huge interest in AIs that can perform tasks on their own.

The problem?

While most of today's apps are capable of understanding the task - including providing a breakdown of how to execute it - they're incapable of actually executing the task.

Mainly because most of today's AIs can't interface with the web.

With Claude 3, however, users will be able to setup truly autonomous GPTs that interact with both the web - and various software tools - via API.

While setting these up could take some serious work, the promise of AIs being able to "work" on their own opens up mind-boggling use cases.

#3 - 99% Recall Accuracy

While many people submit one-and-done prompts, more advanced tasks can require a back-and-forth "conversation" with the AI.

The problem, however, is that many AIs struggle to remember the earlier parts of the conversation (similar to someone with Alzheimer's forgetting what you just spoke about a few minutes ago).

With Claude 3, however, they're achieving near 99% recall accuracy.

And because of that, you'll be able to submit 10, 20 or 100 prompts without the AI getting confused or losing track of what you're working on.

(If you've ever experienced this when conversing with an AI, you know just how frustrating it can be. Making this a highly desirable improvement.)

#4 - Strategy and Data Analysis

In addition to financial forecasting (above), Claude 3 is capable of analyzing charts and graphs, analyzing market trends, data forecasting and more.

From using historical data to build your own investing bot, to forecasting population growth, the ability to work with data and get accurate responses will open up a variety of use cases that just aren't possible with today's AIs.

In conclusion, it appears (the highest level version) of Claude 3 really is the most intelligent AI on the market. But with a focus on business, it's hard to say whether everyday consumers will go out of their way to test its powers.

💡Wrap Up: With pay-per-token pricing, it's clear Anthropic plans to target large companies. But that doesn't mean you shouldn't test it when you have a more demanding task ChatGPT and/or Gemini aren't capable of processing.

🤔Thought-Provoking Question: Would you sign up for pay-per-token pricing instead of a flat fee pay-per-month model?

Interesting Tool ⚙️: While access is limited to certain countries, you can sign up for Claude here.