Meet Google’s most advanced AI yet.

Plus: 🚀 Final 48 Hours – Last Chance before Price Increase!

AuthorAuthor

Tak Lo & Daniel Akpobare
March 10, 2025

Hello and welcome to the Automated, your AI tour guide.

Ever wondered how AI keeps getting better at understanding text?

Google just took it up a notch with Gemini Embedding—its most advanced text embedding model yet. But what sets it apart, and why does it matter? Let’s dive in.

Here’s what we have for you today:

  • 🦾 Gemini Embedding: Google’s Biggest AI Breakthrough Yet.

  • ⚖️ Musk’s Fight Against OpenAI’s For-Profit Shift Isn’t Over Yet!

  • 💻 Prompt engineering for web development.

  • 🛠️ How to build your own AI assistant.

  • 🤖 ChatGPT Prompt Of The Day: Write an email to sell your product.

🦾Gemini Embedding: Google’s Biggest AI Breakthrough Yet.

You see, computers don’t “understand” words the way we do. To them, dog and cat are just random letters—unless those words are converted into something they do understand: numbers.

That’s where text embeddings come in. They transform words, sentences, and even entire paragraphs into dense numerical vectors, allowing AI to grasp meaning, context, and relationships between words.

In short, they help machines read between the lines.

Now, if that already sounds cool, Google just took things up a notch with its latest text embedding model—Gemini Embedding (text-embedding-large-exp-03-07). And let’s just say, it’s shaking things up.

But First—Why Should You Care About Text Embeddings?

Back in the day, conventional language models (aka todays large language models) treated words like isolated islands, completely disconnected from each other.

That meant AI couldn’t understand synonyms, context, or how words actually relate to one another. But text embeddings changed that by positioning words that share meanings or contexts near each other within a multi-dimensional space.

This was a game-changer.

Now, AI can generalize better—even if it hasn’t seen a word before, it can guess its meaning based on context.

It also dramatically improves tasks like document classification, sentiment analysis, and translation while handling massive vocabularies more effectively.

And the best part? It works across multiple languages, like Google’s LaBSE model, which seamlessly processes 109 languages. 

Now that you see why embeddings matter, let’s talk about Google’s latest upgrade—because it’s a big deal.

And Here’s Why:

  • It’s dominating the competition – Gemini Embedding just crushed the MTEB Multilingual leaderboard with a mean score of 68.32—beating the previous best by a whole +5.81 points. That’s like breaking a sprint record and doing it while drinking coffee.

  • It can handle way bigger inputs – This model supports 8K tokens, which is more than double what the previous version could handle. In simpler terms? It can process longer pieces of text without breaking a sweat.

  • It delivers richer, more meaningful embeddings – Instead of the usual dimensions, this model ups the game to 3K-dimensional outputs, meaning AI can understand text with way more depth.

  • It’s smarter about storage – Thanks to a fancy technique called Matryoshka Representation Learning (MRL), you can shrink embeddings down to fit storage constraints without losing accuracy.

And the coolest part? It’s built for the real world. 

Google isn’t just showing off here—Gemini Embedding is designed for practical, real-world use cases. Whether you’re working with:

  • Enterprise search (finding documents buried in a sea of corporate jargon)

  • Finance, legal, and scientific applications (where accuracy is non-negotiable)

  • Or AI-powered recommendations and search (so you actually get useful results instead of junk).

…this model is built to handle it.

And because it supports over 100 languages, it’s perfect for global applications where multilingual AI is essential.

Before you get too excited, Google has made it clear: this model is still in its experimental phase.

That means it’s available via the Gemini API, but it’s not set in stone yet. Also, capacity is limited for now, so don’t expect to throw massive workloads at it just yet.

Google is also still fine-tuning it, so bigger improvements are on the way in the coming months.

The Bottom line?

If AI had a glow-up, this would be it.

Google’s Gemini Embedding isn’t just another incremental upgrade—it’s a huge leap forward in making AI understand and process text in a way that feels more... human.

So, what does this mean for you?

Whether you’re a developer, researcher, or just tired of AI misunderstanding everything, this is a big deal.

Text embeddings are the foundation of modern AI, and with Google supercharging them, let’s just say, the future of AI-powered search, chatbots, and content generation is looking brighter than ever.

Want to learn more? Click here and here.

🚀 Final 48 Hours – Last Chance Before Price Increase!

Hey everyone,

This is your last chance to lock in The Lo Down Premium Experience at the current price before it jumps to $9.99/month!

For just $3.99/month or $39/year (2 months free!), you’ll get:

✅ Weekly AI deep dives – Actionable insights & strategies
🧠 Expert guest columns – Fresh takes from top AI & business minds
🎁 Exclusive Bonus: The AI Insider Playbook ($14.99 value)
💬 1:1 Call Bonus ($500 value): First 10 annual subscribers get a private AI strategy session with me

🔓 NEW 2-Year Deal: $69.99 (Save 30% & get 8 months free!)

⏳ Time is running out! Secure this price before it goes up to $9.99/month. Don’t wait—this offer disappears in 48 hours!

Cheers,

Tak Lo (Editor at The Automated, AI entrepreneur and thought leader. More at thetaklo.com)

⚖️ Musk’s Fight Against OpenAI’s For-Profit Shift Isn’t Over Yet!

Elon Musk just lost the latest round in his legal battle against OpenAI, but the fight is far from over.

A federal judge has denied Musk’s request to halt OpenAI’s transition into a for-profit company—but her ruling raises serious questions about the legality of OpenAI’s move, giving Musk and his supporters a glimmer of hope.

Musk’s lawsuit claims that OpenAI betrayed its original nonprofit mission to develop AI for the benefit of humanity by shifting to a “capped-profit” model in 2019 and now attempting to become a public benefit corporation.

On Tuesday, Judge Yvonne Gonzalez Rogers ruled against Musk’s request for an immediate injunction but pointed out that OpenAI’s nonprofit board stands to gain billions from the conversion, calling it a potential case of “significant and irreparable harm.”

That’s not all—Judge Rogers also noted that OpenAI’s top execs, including CEO Sam Altman and President Greg Brockman, made promises not to use the company to enrich themselves.

Now, the court is prepared to fast-track a full trial in fall 2025 to settle the corporate restructuring dispute.

Musk’s legal team is all in for the expedited trial—but OpenAI hasn’t responded yet.

🚨 What This Means for OpenAI

Legal experts say this decision casts a cloud of regulatory uncertainty over OpenAI’s board.

California and Delaware attorneys general are already investigating the company’s transition, and this ruling could fuel even more scrutiny.

AI safety advocates and watchdog groups are also stepping up pressure, arguing that OpenAI’s shift to a for-profit model could compromise AI safety and put profits ahead of ethical concerns.

However, it wasn’t all bad news for OpenAI.

Judge Rogers ruled that Musk’s legal team didn’t present enough evidence to prove that OpenAI broke a contract by accepting $44 million in donations from Musk before transitioning to a for-profit structure.

She also dismissed claims that Microsoft’s involvement violates antitrust laws and rejected arguments that Musk’s AI startup, xAI, would suffer irreparable harm if OpenAI’s shift moves forward.

🔥 The Bigger Picture

Musk, once a key supporter of OpenAI, has now become one of its biggest challengers.

His AI company, xAI, is directly competing with OpenAI in building cutting-edge AI models, and the legal battle is now intertwined with political power plays under a shifting U.S. administration.

For OpenAI, the stakes couldn’t be higher.

The company reportedly needs to finalize its for-profit conversion by 2026, or some of the billions it recently raised could turn into debt.

Meanwhile, a former OpenAI employee warns that “this transition could threaten public safety”, arguing that OpenAI’s original nonprofit model was meant to ensure AI benefits all of humanity.

If OpenAI goes fully for-profit, they fear nothing will stop the company from prioritizing profits above everything else.

With regulators, investors, and AI safety advocates all keeping a close eye, the coming months will reveal just how many roadblocks OpenAI will face in its quest to become a full-fledged for-profit entity.

Click here to dive deeper into the legal drama!

🧱Around The AI Block

  • Remail: AI-powered email assistant that instantly drafts, replies, and summarizes emails.

  • Hint.fm: AI-driven data visualization tool that creates stunning and interactive visual representations.

  • Postwise: AI tool for generating viral tweets and LinkedIn posts optimized for engagement.

  • Decoherence: AI-powered video generator that creates trippy, music-synced animations from audio input.

  • Genius Sheets: AI tool that automates financial reporting and data analysis in Excel and Google Sheets.

🤖ChatGPT Prompt Of The Day: Write an email to sell your product.

A great sales email doesn’t just inform—it persuades, engages, and converts. Whether you’re launching a new product or reaching out to potential customers, the right words can make all the difference.

With ChatGPT-4, you can craft compelling, high-converting emails that grab attention and drive action.

Here’s a prompt to help you create an irresistible sales email: 

You are an experienced sales copywriter. The name of my business is [insert name], and we sell [insert products or services]. Write a 200-word email that will persuade anyone who reads it to become a customer.

We've Compiled a List of Over 100 ChatGPT Power Prompts.

This should help streamline your interactions with ChatGPT and get the results you need more efficiently.

Best of all, It's free!

That's all we've got for you today.

Did you like today's content? We'd love to hear from you! Please share your thoughts on our content below👇

What do you think of today's email?

Login or Subscribe to participate in polls.

Your feedback means a lot to us and helps improve the quality of our newsletter.

Reply

or to participate.