• BYGEN AI
  • Posts
  • Feeling AI Whiplash? You're Not Alone.

Feeling AI Whiplash? You're Not Alone.

Ever feel like you’re living in two different worlds?

Author

Richard Bohus
June 13, 2025 • Estimated Reading Time: 17 minutes

Outside, life goes on. But online, the news about Artificial Intelligence is a constant storm. It feels impossible to keep up, even if you’re trying to.

One moment, you feel like you have amazing ideas that could change the world. Using great new tools like Claude Code, MCPs, and everything in between feels so amazing. The next, you worry that someone else is already building it faster, or that it will be easily replicated in the near future. Because of this fast output and cheap intelligence, the competition gets brutal. Finding a new moat, a way to protect your work, is extremely difficult. This emotional rollercoaster is something many of us are feeling right now.

The Big Shift: Intelligence is Getting Cheap

So, what’s causing this chaos? It boils down to one powerful idea: intelligence is becoming a commodity.

Like electricity. It used to be a rare, expensive thing. Now, it's everywhere. The same is happening with 'thinking' power.

But here’s the crucial difference: When electricity arrived, it didn’t make us slowly obsolete. It was a tool that augmented our physical abilities, but it didn't challenge our core value. AI is different because it touches the very thing we thought was uniquely human: our minds.

This is a game-changer. It’s no longer about who has the biggest computers, but who has the best ideas to go to market fast.

  • New AI is powerful: Tools like the recently analyzed "o3-pro" are described as a step toward "mass-available superintelligence."

  • Ideas are becoming reality: Projects that were too expensive or difficult yesterday are now possible for small teams or even individuals. That idea you have in your notes? You can probably build a prototype now.

Two Leaders, Two Wildly Different Futures

The confusion gets worse when you listen to the people building the AI. They seem to be telling two completely different stories.

🚨The Alarm: Anthropic's CEO, Dario Amodei, recently warned everyone that AI is coming for skilled jobs, causing widespread anxiety.

☀️ The Dream: On the other hand, OpenAI's CEO, Sam Altman, just wrote about a "Gentle Singularity." He imagines a future where AI helps us solve huge problems and creates so much wealth that society becomes better for everyone.

So, who do you believe? The honest answer is probably somewhere in the middle. Yes, jobs will change. But new opportunities will appear, too.

What This Means for Us

In a world where intelligence is everywhere, what truly stands out? The answer might be simpler than we think.

  1. The "Idea Guy" is King: With AI handling the heavy lifting of coding and complex tasks, your unique ideas and vision become your most valuable asset. But be ready for competition.

  2. Real-Life Connection Wins: As our digital lives get more crowded with AI, real, face-to-face services and local businesses will become more special and valuable. Think of unique experiences that an AI can't replicate.

  3. Superintelligence is on the Horizon: Altman and Sutskever talks about superintelligence by 2030. It’s a wild thought, but it shows how fast this is all moving.

We are standing at a pivotal moment in history. It’s confusing, a little scary, but also filled with incredible potential. The ground is shifting, but for those with creativity and vision, this might just be the most exciting time to be alive.

My "Eureka" Moment with a Real Coding Assistant

I recently decided to take on a project: building a database of the best AI tools by scraping data from the web. Most of the other AI tools databases are not useful at all since most of them just go for the sheer numbers not for the quality. I used to code but had to drop it to focus on my sport career, so my skills were rusty. I’d watched a few videos from the Anthropic team on Claude Code, and it looked intriguing.

I had already tried other AI coding tools like Cursor, Windsurf etc. They’re great, and definitely helpful, but I never had that "eureka!" moment.

Then I tried Claude Code and Claude Desktop (w MCPs). And it was a complete game-changer.

Honestly, I think Claude is the best model out there right now, and the experience is much, much better than anything else I’ve used (for coding). For me, OpenAI's Codex isn't nearly this useful or I haven’t got the gist of it.

The real superpowers come from using the Claude Desktop app with MCPs. Giving it access to your computer's files and data makes it incredibly intuitive. It stops being just a code generator and starts acting like a true assistant that does things for you.

For example, my downloads folder was a complete mess. I asked Claude to clean it up, and it sorted everything into new folders based on the content of the files. It just did it.

When it comes to my scraping project, the process is so smooth it feels like I’m handing tasks off to a human assistant. Using libraries like Smithery's MCP library has been a huge help as well. It handles the work seamlessly.

The only annoying part is the rate limiting, which can slow you down. But that’s a small price to pay. For the first time, I feel like I have an agent that can truly take on and complete tasks from start to finish.

BYGEN Updates

Alright, here's what's currently on my radar at BYGEN:

  • AI News Aggregator Project: It’s coming along nicely, the backend is functioning and we are putting together the design right now. Later we want to implement agents that always watching the trends for you no matter what.

  • Custom Midjourney Prompter: I will share my new custom GPT in the next newsletter so you guys can try it out. I use it for almost all of my artworks so can’t wait to give it to you.

  • Website Development Progress: It’s mostly done, I want to add the ai tools database that I talked about and other handy AI tools that I found on the internet.

📡 AI News worth knowing

OPENAI

🥊 OpenAI Strikes Back: Unveils Strategic O3 Pro & Slashes Prices

Just when it seemed like Google and Claude had dethroned the king, OpenAI is roaring back into the game. With a massive 80% price cut on their standard o3 model and the release of the incredibly powerful o3 Pro, they've got back on top, creating a formidable one-two punch in the AI race.

TLDR: After facing stiff competition, OpenAI has returned to the lead with a dual strategy: a huge price cut on the standard o3 model and the launch of o3 Pro, a new AI that acts like a strategic partner. o3 Pro's power requires thoughtful handling and is best reserved for complex problems, not simple queries.

Key Features of o3 Pro

  • Resonant Intelligence: o3 Pro delivers insights that are not just accurate but deeply resonant, feeling personally tailored to the user.

  • Strategic Superiority: It has shown a superior ability to handle high-level strategic thinking, even knowing when to stop and admit its limitations.

  • Beyond Benchmarks: Standard performance benchmarks don't fully capture the qualitative "feeling" of how good this model is at genuine strategic reasoning.

  • Fact-Checking is Crucial: Its responses are so persuasive that it's highly recommended to cross-check its outputs for accuracy.

How to Get the Most Out of O3 Pro

  • Choose the Right Tool for the Job: Now more than ever, knowing your goal is crucial to select the right model. o3 Pro is built for depth, not speed.

  • Reserve it for Hard Problems: This model is not for casual chats. It can take up to four minutes to process a simple "hello" because it's designed for deep, deliberate thinking. Use this model for really hard problems and give it as much context as possible to get the best answer.

Source: @benhylak

Why it matters

OpenAI's new lineup signals a shift in how we must approach AI tools. It’s no longer about which model is "best" overall, but which is best for a specific task. With a cheaper, faster option in the standard o3 and a deep-thinking partner in o3 Pro, the skill of choosing and prompting the right model is now more valuable than ever.

Meta

🤖 Meta Goes All-In on AI with $14.3B Investment in Scale AI

Source: Bloomberg

TLDR

Frustrated by sluggish internal progress, Mark Zuckerberg is radically overhauling Meta's AI division. The company is planning a major restructuring of its AI teams while simultaneously making a colossal $14.3 billion investment for a 49% stake in Scale AI, a key player in providing data for training AI models. The deal includes recruiting Scale AI's CEO, Alexandr Wang, to lead a new "superintelligence" unit at Meta.

Key Details:

  • Internal Overhaul: The move comes amidst Zuckerberg's frustration with the company's slow AI progress, particularly the performance of its Llama models, leading to a major shake-up and planned cuts of existing AI teams. He is reportedly taking a more direct, hands-on approach to micromanage AI ventures.

  • Massive Investment: Meta is investing $14.3 billion for a 49% stake in Scale AI.

  • Strategic Hire: Scale AI's CEO, Alexandr Wang, is being brought in to lead a new, elite "superintelligence" group aimed at accelerating Meta's push toward artificial general intelligence (AGI).

  • Performance-Based Cuts: This restructuring follows earlier layoffs in 2025 framed as removing "low performers" to increase overall efficiency and focus resources on AI.

Why it matters:

This isn't just an investment; it's a high-stakes pivot. The internal team cuts and the massive external investment are two sides of the same coin, revealing deep dissatisfaction with the current pace of innovation at Meta. By bringing in top talent like Wang and pouring billions into a strategic partner, Zuckerberg is signaling he is willing to dismantle and rebuild his AI operations to compete directly with rivals like Google and OpenAI. The strategy is aggressive: tear down internal silos, inject new leadership, and spend heavily to secure a leading role in the future of AI.

Creative AI News

Midjourney

⚖️ Disney and Universal Sue Midjourney, Sparking a High-Stakes AI Copyright Battle

TLDR:

In a landmark move, entertainment giants Disney and Universal have filed a lawsuit against the popular AI image generator Midjourney, accusing it of "massive copyright violation." It’s hard to argue since in the document the examples are crystal clear. This legal battle could set a critical precedent for how AI companies are allowed to use copyrighted material for training their models, potentially reshaping the future of generative AI.

The 110 pages lawsuit, filed in a California district court, alleges that Midjourney has illegally used "countless" copyrighted works to train its AI. The studios claim that the platform functions as a "bottomless pit of plagiarism," allowing users to generate images of iconic characters like Darth Vader, Elsa, and Homer Simpson.

“By helping itself to Plaintiffs’ copyrighted works, and then distributing images (and soon videos) that blatantly incorporate and copy Disney’s and Universal’s famous characters, without investing a penny in their creation - Midjourney is the quintessential copyright free-rider and a bottomless pit of plagiarism,” the complaint reads. 

Key Details of the Lawsuit:

  • Allegation of Infringement: Disney and Universal contend that Midjourney built its business by unlawfully scraping and using their intellectual property without permission or compensation.

  • Requested Action: The lawsuit seeks to force Midjourney to stop using their copyrighted works and implement measures to prevent the generation of infringing images. It also asks for unspecified damages.

  • Industry-Wide Implications: This is the first time major Hollywood studios have taken legal action against a generative AI company, signaling a significant escalation in the fight over copyrighted training data.

Why It Matters:

This case will rigorously test the "fair use" argument that many AI companies have relied on, which posits that scraping public data for training purposes is legally permissible. A victory for the studios could have a domino effect across the AI industry.

Some believe Disney targeted Midjourney because it is a smaller, self-funded company, making it an easier opponent than industry giants like OpenAI. As a bootstrapped project, Midjourney lacks the extensive legal and financial resources of venture-backed competitors. A loss for Midjourney could embolden other copyright holders to file similar lawsuits against other AI developers, potentially leading to a wave of litigation that could stifle innovation or force a fundamental change in how AI models are built and trained.

AI Video

🎬The Future of Ads Just Aired: How an "Unhinged" AI Ad Rocked the NBA Finals

Source: PJ Ace

If you were glued to the NBA Finals this week, you might have seen it: a wild, fast-paced, and frankly "unhinged" commercial for the trading platform Kalshi. It was a memorable 30 seconds of television.

But the real story isn't just the ad itself. It's how it was made.

My good friend PJ Ace created the entire spot in just a few days, using a suite of AI tools. I had the privilage to work with him on the Kling 2.0 trailer and his workflow and vision on creating viral content is amazing. Check out his newsletter where he spills the BTS on how he makes these.

So, how did a high-energy, national commercial get made faster than it takes to get a new sofa delivered? PJ broke down his groundbreaking process.

  • The Idea: Start with a rough script and a zany, over-the-top premise.

  • The AI Brains: Use Gemini and ChatGPT to brainstorm and transform the script into a detailed list of shots and powerful text prompts.

  • The AI Director: Generate video clips using Google Flow, a platform powered by Veo 3—Google's state-of-the-art video model that's turning heads in the creator community.

  • The Human Touch: Edit the AI-generated clips together using traditional software like CapCut and Adobe Premiere to polish the final product.

Why It Worked: Leaning Into the Weird

We all know AI can be… unpredictable. Instead of fighting it, PJ embraced it.

The ad's chaotic, quirky style was a deliberate choice that played to the strengths (and current oddities) of AI video generation. By building the concept around the AI's "weirdness," he created something that felt fresh, unexpected, and perfectly aligned with the tagline: "The world's gone mad. Trade it."

What It Means for Creators: A 95% Cost Reduction

This is the number that's shaking up the industry.

Ace claims he created the Kalshi spot with a 95% cost reduction compared to a traditional ad campaign. By replacing massive crews, expensive locations, and lengthy production schedules with AI-powered tools, he collapsed the process into just two or three days.

The takeaway? The barriers to entry for creating high-quality video are evaporating. As the technical side of production gets easier, what becomes truly valuable are the skills AI can't replicate:

  • Taste & Vision: Knowing what makes a compelling story.

  • Creativity & Originality: Coming up with unique ideas that cut through the noise.

The Kalshi ad is more than just a clever commercial; it's a proof-of-concept for a new era of filmmaking. The tools are here, and as PJ Ace has shown, the creative revolution is already being televised.

I haven’t explored the new Veo 3 since I’m based in Europe, but now it’s available on Freepik so I will take it for a spin.

Other News

QUICK HITS

Trending tools

  • Guidde is the generative AI platform for business that helps your team create video documentation 11x faster.

  • Gumloop - Best AI automation in the game for enterprise

  • Dia - The Browser Company’s new AI-first web browser. They should have just keep building Arc. I will dive in and test it.

  • RunnerH - Agentic orchestration tool.

See you soon,

R from BYGEN

Reply

or to participate.