This week, we help readers figure out which AI subscription is best for them, and which they should cut. Then, we look OpenAI’s aggressive spending spree, including investments in a brain computer interface company. Let’s dive in ⬇️

In today’s newsletter ↓
📬 Choosing the best AI subscription in 2026
🏗️ OpenAI seeds future with hardware and neural tech
💊 NVIDIA accelerates AI powered medical breakthroughs
🫧 McKinsey automates graduate screening with chatbots
🧩 Weekly Challenge: AI model showdown

The $20-ish monthly AI subscription fee has evolved into a comprehensive suite of services. While model intelligence is a major factor, the peripheral benefits like cloud storage, specialized search tools, and software integrations often justify the cost.

Power users are moving away from a one-size-fits-all mentality and selecting tools that integrate natively into their professional workflows. Choosing the right plan requires looking past benchmark scores and auditing the actual features, privacy defaults, and context windows included in each ecosystem.

The Google AI Pro plan costs $19.99 monthly and is currently the most aggressive bundle in the market for data-heavy users.

If you live in Google Workspace and need an assistant that can synthesize massive amounts of data across your entire digital life, this is your winner. Also, since Gemini offers so many features across Google products, this subscription definitely gives you the most bang for your buck.

Claude Pro by Anthropic costs $17 monthly and focuses its value on high-performance professional workflows and nuanced communication.

This plan is the perfect choice for the precision-oriented professional who values high-quality writing and deep technical execution above all else. However, if there is a downside, it’s cost. Compared to other SOTA AI companies, Claude offers the least for the most.

ChatGPT Plus costs $20 monthly and provides the most feature-dense option for generalists who want a "Swiss Army Knife" for every task. This subscription is ideal for most people from the curious generalist to the master jack of all trades who wants a versatile creative partner and access to the widest variety of specialized AI agents.

ChatGPT has generally slower response times than most other leading models, isn’t as specialized as Claude, and not as embedded in an ecosystem as Google, but it still holds its weight.

Perplexity Pro costs $17 monthly and is the premier choice for professionals who rely on verifiable citations and academic search.

Dedicated researchers and fact-checkers who demand verifiable citations and real-time search accuracy will find the most value in this plan. However, many long term users observe that its search isn’t necessarily leaps and bounds better than Gemini or other leading models.

The quest for AGI has moved into physical infrastructure and biological interfaces. OpenAI’s long-term roadmap now includes two massive investments aimed at fundamentally changing our relationship with technology.

The company recently confirmed a $10 billion partnership with Cerebras Systems, a chipmaker building the world’s largest AI processors. Unlike traditional clusters of thousands of small chips, Cerebras builds purpose-built AI systems using a single wafer-scale engine.

This architecture eliminates data bottlenecks by housing massive compute and memory on one piece of silicon, specifically targeting the low-latency inference required for real-time agents.

Reports show OpenAI will buy compute capacity from the startup to ensure its chipmaker can power ChatGPT more efficiently. For users, this means AI that responds instantly, enabling natural, instant interactions that were previously impossible due to lag.

While Cerebras addresses the machine, OpenAI’s role in Merge Labs’ seed round focuses on the human. Co-founded by Sam Altman, Merge Labs aims to bridge biological and artificial intelligence via high-bandwidth brain-computer interfaces (BCIs).

OpenAI is investing in the brain-tech startup because advanced computing requires more direct interfaces. By allowing humans to express intent more efficiently, technology becomes vastly more powerful. Merge Labs is developing approaches to BCIs that combine biology with AI operating systems designed to interpret noisy neural signals.

This two-pronged strategy shows OpenAI is thinking decades ahead. They are securing massive compute for next-gen models while building a neural operating system for seamless human interaction. AI will be central to Merge Labs’ research, accelerating progress in bioengineering and neuroscience.

Ultimately, these investments in brain-computer interfaces suggest a future where AI isn't just a tool on a screen, but a high-speed extension of human agency. These physical bets ensure that future intelligence is supported by infrastructure capable of keeping pace with our growing digital ambitions.

🧬 NVIDIA teams with Lilly to reinvent drug discovery

🚶 Founder exodus hits Mira Murati’s startup as top talent flees back to OpenAI

🧩 Gemini enters its hyper personal era for billions of users

⚠️ Warning issued as photo realistic AI fuels an explosion of harmful digital content

⚡ White House proposes emergency power auctions to fund AI data center growth

💼 McKinsey forces job seekers to collaborate with AI bots during final interviews

Here is how to clean it up:

🥇 The Creative Stress Test: Take a complex writing prompt and run it through the free versions of ChatGPT, Claude 4.5, and Gemini 3. Look for which one captures your specific tone and nuance without you having to over-prompt it.

🥈 The Document Deep Dive: Upload a 50+ page PDF to each model and ask for a summary of a very specific, obscure detail buried on page 37. Note which model hallucinated and which one found the exact data within seconds.

🥉 The Integration Audit: Look at where you spend 80% of your work time. If it’s in Google Sheets and Drive, spend an hour using Gemini. If it’s in coding editors or local files, give Claude 4.5’s desktop app a spin.

🏅 The Financial Verdict: Based on the three steps above, cancel the subscriptions that aren't saving you at least 30 minutes a day. In 2026, your AI should be a profit center, not just a monthly expense.

Are you experiencing AI subscription burn out yet? And, will OpenAI’s latest acquisitions and investments pay off in the long run? Hit reply and let us know your thoughts.

Zoe from Overclocked