Did DeepSeek Steal its AI Secret Sauce Again?!

Plus, master the art of beating the ATS with keyword precision

In partnership with

Welcome, Automaters!

DeepSeek is back in the headlines—and this time, it feels way too familiar. They just dropped a new model that’s crushing math and coding benchmarks, but the AI world isn’t exactly convinced.

The big question is: Did they really train it the hard way… or are they borrowing someone else’s brainpower again?

Here's what we have for you today

 😱 AI Copy Culture Hits a New Peak

China’s AI lab DeepSeek just rolled out a new reasoning model—R1-0528—and it’s absolutely crushing math and coding benchmarks.

But here’s the thing: no one really knows what data they used to train it… and that’s where the drama kicks in.

A Melbourne-based developer, Sam Paech—who builds emotional intelligence tests for AI —thinks he’s cracked the case.

According to Paech, DeepSeek’s model seems to favor the same words and phrasing as Google’s Gemini 2.5 Pro.

And that alone isn’t a smoking gun. But another developer — the pseudonymous creator of a “free speech eval” tool for AI called SpeechMap — noted that the “thoughts” DeepSeek generates while reasoning through a task read like Gemini traces.

Coincidence? Eh, unlikely.

Taken together, they suspect the model may have been trained on Gemini outputs. And if this whole thing feels a little too familiar, that’s because DeepSeek’s kinda been caught in similar situations before:

  • Back in December, another DeepSeek model randomly identified itself as ChatGPT. Yep—“I copied your homework and forgot to change the name” energy.

  • Earlier this year, OpenAI and Microsoft flagged suspicious data scraping activity from OpenAI developer accounts—accounts they believe were tied to DeepSeek.

  • And the method behind all this? Distillation—aka, training your AI by siphoning knowledge from a more powerful model. It’s not illegal, but it’s definitely against OpenAI’s terms if you’re using it to build a competing model.

Now, to be fair: the internet is one big AI soup right now. There’s so much machine-generated junk floating around, it’s getting harder and harder to find clean data.

In fact, things have gotten so messy that models are starting to sound the same—just because they’ve all been trained on the same AI-spammy content. But still… DeepSeek’s case feels a little too fishy to ignore. 

So, how’s the AI world responding to all this distillation drama?

They’re tightening up—fast:

  • OpenAI now requires government-issued ID to access certain advanced tools. And spoiler: China’s not on the list.

  • Google is now scrambling Gemini’s output trails to make them harder to scrape.

  • Anthropic is doing the same with Claude, all in the name of protecting their competitive edge.

The real takeaway?

DeepSeek’s model might be smart—but the AI world’s not totally buying that it got there on its own. In fact, everyone’s wondering if it’s really that smart—or if it just passed the test using someone else’s notes.

You’re doing breakfast wrong

Let’s face it—most breakfast options just don’t cut it.

Toast? Too light. Cereal? Mostly sugar. Skipping it altogether? Not ideal.

If you want real fuel to power your day, it’s time to upgrade to Huel Black Edition. This ready-in-seconds shake is packed with 40g of plant-based protein, 27 essential vitamins & minerals, and 0 artificial sweeteners—just science-backed nutrition to support your muscles, digestion, and more.

Oh, and did we mention? It’s delicious.

Right now, first-time customers get 15% off, plus a free t-shirt and shaker with code HUELSPRING, for orders over $75.

🎬 OpenAI Firing Fiasco Gets the Hollywood Treatment

In case you forgot (or just repressed it), back in late 2023, OpenAI's board fired Sam Altman—yes, the poster boy of AI—only to rehire him five days later after a whirlwind of internal chaos, public backlash, and (we’re assuming) some very dramatic Slack messages.

Well guess what!

That entire saga is officially being turned into a movie titled “Artificial.” 

According to The Hollywood Reporter, it’s currently in development over at Amazon MGM Studios—and here’s what we know so far:

  • Luca Guadagnino (aka the visionary behind Call Me by Your Name and Challengers) is in talks to direct. So expect it to be visually stunning, even if it’s about boardrooms and boardroom betrayals.

  • Andrew Garfield might take on the role of Sam Altman. I mean, if anyone can pull off “boy genius caught in a corporate meltdown,” it’s probably him.

  • Monica Barbaro, (a complete unknown) is reportedly up to play Mira Murati (OpenAI’s then-CTO, who briefly took the wheel after Sam got booted).

  • Yura Borisov (from Anora) could be cast as Ilya Sutskever—the OpenAI co-founder who helped remove Altman... and then very publicly regretted it (Talk about character development.)

  • And here’s the twist: Simon Rich, a writer from Saturday Night Live, wrote the screenplay. So yes, this might be the first AI boardroom thriller with punchlines and comedic moments—because let’s face it, AI drama already feels like a satire.

And come on, are we even surprised this is headed for the big screen?

Curiosity about OpenAI, Altman, the boardroom saga, and the power dynamics behind the world’s biggest AI company is only growing fast.

The internet’s been swirling with rumors, theories, and “what really happened” takes—so maybe, just maybe, this film could shed some light… or add more mystery 👀.

So if you missed the whole drama, still don’t get why everyone was freaking out, or you’re just AI-curious... Then hey—this film might be your popcorn-fueled crash course in OpenAI chaos.

Just maybe don’t take it all as gospel, because let’s be real: Hollywood loves a good publicity stunt.

As for when we’ll see it? There’s no official timeline yet. But reports say Amazon’s aiming to kick off production this summer—so we’ll probably be waiting a bit.

But until then? We’ll have to keep guessing, theorizing, and watching ChatGPT updates like it’s a reality show.

Here’s the full story.

✍️ Editor’s Corner

Hey Automaters,

I hope you’ve enjoyed the AI WoD trailers and videos that we’ve been making, as much as we have had making them.

They may seem like easy videos to make, but it’s taken almost a year of planning and testing with different lighting and products. So what you see now is a product of real deep work and iterations to perfection.

And if you haven’t noticed (though some of you did!) - you can now buy daily AI WoDs for just $1.99! Just click the buy now link or here and you have today’s WoD delivered into your inbox!

Cheers

Tak Lo

🧱Around The AI Block

🤖 ChatGPT Prompt Of The Day: Optimize Your Resume for ATS

Your resume might be impressive, but if it doesn’t speak the language of Applicant Tracking Systems (ATS), it could be overlooked before a human ever sees it.

This prompt is designed to help you fine-tune your resume so it aligns perfectly with the language and requirements of a job description. By identifying missing keywords and strengthening your phrasing, you'll increase your chances of making it through the automated screeners and landing an interview.

Here’s How to Use This Prompt Effectively:

  1. Be specific. Paste both the job description and your resume in full. The more detail you provide, the better the optimization.

  2. Target one role at a time. This prompt works best when tailored to a single job description so the keyword matching is spot-on.

  3. Ask for rewrites. Don’t just stop at keyword suggestions — ask for help rewriting bullet points or sections to naturally include those keywords.

  4. Keep your tone in mind. Make sure the rewrites still reflect your voice, professionalism, and readability.

Here’s an example:

Act as an ATS keyword optimization expert. I will provide my current resume and a specific job description. Your task is to analyze both and identify missing keywords or qualifications that may cause my resume to get filtered out by AI screeners.

Perform a side-by-side comparison between the job description and my resume. Highlight any missing or weak keyword matches and recommend specific phrases or bullet points to include.

Maintain a natural tone while boosting alignment with the job. Provide:

A revised version of the relevant resume sections (or full resume if needed)

A brief explanation of your keyword strategy and how it improves ATS performance.

Inputs:
Job Description: [Paste the full job posting]
My Resume: [Paste your resume text]

Here’s a sneak peek:

Upgrade now to see this whole month’s prompt videos and more, or buy TODAY’S WOD for just $1.99

Is this your AI Workout of the Week (WoW)? Cast your vote!

Login or Subscribe to participate in polls.

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'd 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.

🚀 Want your daily AI workout?

Premium members get daily video prompts, premium newsletter, an no-ad experience - and more!

Already a paying subscriber? Sign In.

Premium members get::

  • • 👨🏻‍🏫 A 30% discount on the AI Education Library (a $600 value - and counting!)
  • • 📽️ Get the daily AI WoD (a $29.99 value!)
  • • ✅ Priority help with AI Troubleshooter
  • • ✅ Thursday premium newsletter
  • • ✅ No ad experience
  • • ✅ and more....

Reply

or to participate.