Gemini 4 Leaks: What They Show, What Is Unverified, and Why Google's Data Matters

Dan Draguta — Sun, 04 Oct 2026 05:31:24 +0000

Important up front: none of what follows is confirmed by Google. Everything here is leaked, rumoured or posted by third parties. Treat it as speculation, not news.

Why I made this video

Most "Gemini 4 leaked" videos are polished and scripted. I wanted to do the opposite: open a browser, look at the leaks as a normal user would, and react live. No script, no strategy.

Before we start, one question to keep in mind. Every AI model needs data to produce a result. Who has the most data right now? Google. That, more than any benchmark, is the angle that could change things. Google either makes a serious comeback with workflows and automation built on that data, or this becomes another flop. Either outcome would surprise nobody.

Step one: ask Gemini itself

I asked Gemini directly whether the Gemini 4 leaks are real. The short answer it gave: no, they are not officially confirmed, and parts of them appear fabricated or unverified. That sets the tone. From here on we are speculating.

The "Argon" checkpoint claims

One of the first accounts to post about it on X shared what it described as the first output from a Gemini 4 Pro checkpoint, internally code-named Argon: a block-built 3D scene that looks like something you would make in Blender. The claims attached to it:

The output took about 2.4 minutes on high thinking effort.
An output limit of around 256K tokens, compared with 64K in previous Gemini models.
A context window of 2 million tokens at launch, roughly double what competitors offer today.

If those numbers hold, more context means much bigger jobs in one go: whole codebases, long document sets, full projects in a single request.

Benchmarks and pricing

Several posts claim wild benchmark scores, high speed and low pricing. The figures circulating are roughly $2.25 per million input tokens and $11.25 per million output tokens. Nobody has verified the benchmarks or the prices. If the pricing is real, it would put pressure on the whole market.

The 3D demos

This is where the leaks get interesting. The examples shared on Reddit and X include:

A detailed race car
A 3D city building, so architecture as well as objects
A ship, and transformations of scenes
The classic "pelican on a bicycle" test, with customisation
A pagoda shown side by side: the older Flash model's version next to the claimed Gemini 4 Pro version. The difference is night and day.

If any of these came from one prompt with no outside software, that is a big shift. Anyone who has tried to get results like these in Blender or another 3D tool knows how much work they take.

What it would mean if it is real

Websites will not stay the same. Interactive 3D that used to need a specialist could become a prompt.
Workflows and automation. Huge context plus Google's data and products is where the real leverage would be.
Price pressure. Cheap, capable models change what is worth building yourself.

My take

We have to wait for the official release. But there is rarely smoke without fire, and if the final model looks anything like these leaks, it will probably surprise us even more.

The lesson for your business is the same either way: don't build on any single model. Build your own data, workflows and stack so you can switch to whichever model wins.

Want an AI setup you own?

I build AI workflows and automation that you own, so you can switch models without starting over. Book a free call and we will look at what makes sense for your business.

Originally published at wolfitway.com.

Stop Copying Successful Businesses: Build a Sovereign Moat Instead

Dan Draguta — Sat, 03 Oct 2026 20:01:40 +0000

Stop trying to copy successful businesses. It doesn't matter if you have their exact funnel, their exact ad spend or their exact code.

You will still lose. The copy-paste model of entrepreneurship is broken, and in the age of AI the only thing that can't be copied is you.

The copycat paradox

The usual advice is: find a competitor, reverse-engineer their success and do it better. That is advice for employees, not for founders who want to own what they build.

We are moving from the information age to the synthesis age. In the synthesis age the how is free. The why is the only asset you have.

The industrial mindset is holding you back

We were taught that a business is a machine: inputs go in, outputs come out, and if you want the same output you copy the machine. That worked in 1950 and in 1990.

Today the machine is software, and software can be copied endlessly. If your business depends on a specific stack, a specific marketing tactic or a specific pricing model, you are a commodity. Commodities compete on price, and that is a race to the bottom.

The clone focuses on the what and the how: "Which feature don't they have? How do I build a cheaper version?"
The sovereign focuses on the who and the why: "Who am I uniquely positioned to serve, and why am I the only logical choice?"

You can copy the map, not the territory

A business isn't a product. It is a context engine. When you look at Apple or Tesla you see the product. You don't see 20 years of context: the relationships, the supply-chain quirks, the cultural timing that let that product exist.

Copying a business means copying the map while trying to navigate a territory that has changed since the map was drawn. In tech, tactics last about six months. Spend a year copying a competitor's strategy and by launch it no longer works. You are chasing a ghost.

The contextual moat: a three-part system

Data gravity

Data isn't just numbers. It is gravity: it pulls people in. A clone has generic data ("here is a generic guide to losing weight"). A sovereign has proprietary data ("here is three years of biometric data from my community of 1,000 users showing a link between sleep and cortisol").

You can't copy that data. You have to generate it and earn it. The business isn't the software. It is the feedback loop that makes the software smarter every day.

Network topology

This is who you know and how you connect them. A clone builds a user base. A sovereign builds a network.

A SaaS tool with only users will churn. A SaaS tool that connects specific freelancers with specific high-end clients isn't selling software any more. It is selling access, and access can't be copied.

A real example: someone I know runs a niche job board for underwater welders. Anyone could copy the website. Nobody can copy the fact that he personally knows the top 50 hiring managers in that industry. That is his moat, and he earned it.

3. Sovereign identity

This is the hard part for tech people: you have to put yourself into the business. When AI can write 1,000 blog posts a minute, the only thing that matters is trust. Trust doesn't scale in the usual way. It is earned through friction.

When you copy someone, you are a guest in their house, playing by their rules. When you build your own moat, you set the rules: "This is how we do things here. If you don't like it, that's fine." That friction is a filter. It builds a following, not just a customer list.

The audit: do this now

Look at your business model and product and ask one question:

If a competitor stole my entire codebase and content library tomorrow, would I still win?

If the answer is no, you are in the copycat trap. If it is yes, it's because at least one of the three is present: data, network or identity.

From features to philosophy

Stop selling 10 GB of storage. Start selling ownership of your data. Stop selling "marketing services". Start selling a contrarian growth philosophy, for example one that rejects paid ads.

Yes, this alienates people. That is the point. You lose the people who would have copied you anyway and attract the people who share your worldview. You can't be independent while begging for attention by copying trends.

You become sovereign when you become the source, and when you are unapologetic about who you are, what you want and what you do.

The moat nobody talks about: shared struggle

The biggest moat isn't the code or the marketing. It is the shared struggle. When you copy a business you inherit its success but not its scars, and your audience connects with your scars: the time you failed, pivoted or nearly went broke.

A clone has no scars. It is smooth, polished and boring. So document the war, not just the victory. If you are building a SaaS, show the bugs, the server crashes and the customer complaints. That transparency is the ultimate copy protection, because a copycat can't fake your pain.

The bottom line

You can't copy a business because you can't copy the context, the timing or the messy network of human relationships that made it work. Stop looking at your competitors and look at your own architecture:

Build data gravity.
Build a network.
Build a sovereign identity.
Embrace the struggle. That is your moat.

Don't be a clone. Be the source.

Build your moat with me

I help founders use AI to build moats like these instead of letting AI wipe them out: your own data, your own stack, no vendor lock-in. Book a free call and we will map out yours.