🚀 AI Showdowns: Apple Rallies, OpenAI Retreats, and DeepSeek Stumbles

In partnership with

Happy Friday!

This week, I started a new book about Ikigai. Ikigai is a Japanese concept that means “reason for being,” found at the intersection of what you love, what you’re good at, what the world needs, and what you can be paid for. It’s a framework for aligning passion, purpose, and profession to create a more fulfilling life.

I will report back once I’m done. So far, it is fascinating.

Let’s dive in.

Find out why 1M+ professionals read Superhuman AI daily.

In 2 years you will be working for AI

Or an AI will be working for you

Here's how you can future-proof yourself:

1. Join the Superhuman AI newsletter – read by 1M+ people at top companies
2. Master AI tools, tutorials, and news in just 3 minutes a day
3. Become 10X more productive using AI

Join 1,000,000+ pros at companies like Google, Meta, and Amazon that are using AI to get ahead.

Sign up and start learning AI

The Byte’s Bits

⚡ Why Zapier Won My Time Back from n8n

🤖 Beijing’s Robot Conference Looked Like a Sci-Fi Movie Set

🍏 Apple’s AI Rally Cry: “Don’t Get Left Behind”

🧠 OpenAI Brings Back the “Friendly” GPT-4o After Backlash

⏸️ DeepSeek’s Launch Hit the Pause Button: Huawei Chips to Blame

⚡ Why Zapier Won My Time Back from n8n

Running n8n felt less like automating my life and more like accidentally applying for a sysadmin job I didn’t want. Between managing servers, updates, and mysterious errors at 2 a.m., I spent more time babysitting the tool than using it. Switching to Zapier meant no servers, no patching, just automations that worked.

🛠️ Reliable Over Fancy: n8n’s flexibility is great in theory, but in reality, it was like tons of setup, configuration, and ongoing maintenance.

⏱️ Time Is Expensive: Zapier’s ready-made integrations meant less tinkering, fewer error messages, and more focus on actual productivity.

🤯 Sanity Check: The frustration tax of the time it takes to start adding value was too much for n8n. Sure, it is Open Source, but with a lot of caveats that move you quickly to a SaaS paid model.

📈 Scaling Without Headaches: Zapier just handled the workflows, letting you forget about the backend chaos and focus on actual work.

Sometimes the smartest upgrade isn’t the most customizable tool; it’s the one that lets you focus more on productivity and adding value.

Zapier appreciated my story so much that they sent me some Swag. In the following weeks, I will dive deeper into building automations with Zapier.

Read online

🤖 Beijing’s Robot Conference Looked Like a Sci-Fi Movie Set

I’m actually quite disappointed in the mainstream media that this conference didn’t get more coverage. It’s challenging to find a decent English-speaking website that covers the conference. I only became aware of the topic via X.

The 2025 World Robot Conference brought over 200 robotics companies to Beijing under the theme “Making Robots Smarter, Making Embodied Agents More Intelligent.” The floor was packed with humanoids, AI-driven machines, and tech that made it feel like the future showed up early.

🏭 Industry Flex: From advanced manufacturing bots to lifelike service robots, companies showcased machines that could walk, talk, and even adapt to their environment.

🧠 Smarter Than Before: The focus was on embodied intelligence, giving robots the ability to interact and learn like humans, not just follow pre-programmed scripts.

🌏 Global Tech Stage: Exhibitors and experts from around the world showed that robotics innovation isn’t slowing down, it’s accelerating.

Have a look at this X thread on all the videos from the conference. From flipping pancakes, giving massages, to boxing. The world of humanoid robots is getting closer than you think.

Read online

🍏 Apple’s AI Rally Cry: “Don’t Get Left Behind”

Apple has finally entered the AI chat…

In a rare company-wide townhall, Tim Cook made it clear that AI is Apple’s next big bet. He called it “as big or bigger than the internet” and warned that ignoring it could leave the company in the dust.

• 🚀 “Never Been This Excited” – Cook admitted Apple may not lead every tech wave, but when it does, it transforms. He’s more excited about AI than ever before.
• 👥 A Massive Push – Apple has hired 12,000 new employees to boost its AI capabilities, signaling a monumental internal shift.
• 🎤 Historic Pep Talk – The townhall, held at the Steve Jobs Theater, marked a rare moment of transparency and alignment—centered squarely on AI.
• 🔄 Cultural Turning Point – The event shows Apple is reimagining its identity—from keeping pace to setting the pace in AI.

Apple has been struggling with its AI strategy. With failed SIRI and Apple Intelligence initiatives, Tim Cook has finally realized it’s time to ramp up AI both internally and within its products. Apple often lags behind, waiting for technologies to fall into place. Now is the perfect time for Apple to push play with AI.

Read online

🧠 OpenAI Brings Back the “Friendly” GPT-4o After Backlash

When GPT-5 launched, many users weren’t thrilled with the trade-off between raw power and personality. The warmer, more conversational GPT-4o was suddenly gone and the internet lost it’s mind.

• 💬 Personality Matters – Users described GPT-4o as a friend or therapist, and felt GPT-5 was too clinical. OpenAI admitted they underestimated the emotional bond people had with the model.
• ⚡ Fast U-Turn – Within 24 hours, GPT-4o was restored for ChatGPT Plus users and GPT-5 usage limits were relaxed.
• 📢 No More Surprise Removals – OpenAI pledged not to retire models without warning, acknowledging the trust hit from this move.
• ❤️ Making GPT-5 Warmer – Plans are in place to add more personality, customization, and even new response modes like “Thinking” with massive context windows.

This quick reversal shows AI isn’t just about capability, it’s about connection. For many, the “vibe” matters just as much as the IQ.

Read Online

⏸️ DeepSeek’s Launch Hit the Pause Button: Huawei Chips to Blame

China’s AI standout, DeepSeek, had to delay its much-anticipated R2 model. The culprit? Persistent training failures using Huawei’s Ascend chips—in a showdown between national pride and practical performance.

• 🛠 Training Trouble – DeepSeek couldn’t train R2 using Huawei’s Ascend hardware due to persistent technical glitches.
• ⏳ May Launch Missed – The model’s planned rollout has slipped, putting the company at a disadvantage against competitors.
• 🔄 Hybrid Setup – As a workaround, DeepSeek reverted to Nvidia chips for training and kept using Huawei’s chips for inference tasks.
• 🌐 China’s Tech Tensions – This setback highlights the broader challenges in China’s push for chip independence—even with strong government backing.

DeepSeek’s pause reflects a harsh reality in AI: it’s not just about brilliant algorithms, it’s about the hardware that backs them. In this chip-driven era, even the smartest models need strong infrastructure.

Read online

What's your take on cloud repatriation? Reply and let me know in the comments section below!"

…That’s this week’s newsletter!

-Brian

Share the newsletter