The Multiverse Next Door: How Quantum Computing is Redefining Reality

Hey there,

I can't believe we're already in 2025! While everyone's still working on their New Year's resolutions, let me catch you up on some fascinating developments from the past few weeks that are shaping our AI future.

Last weeks were particularly exciting - Google dropped a mind-bending announcement about their quantum chip potentially proving parallel universes exist (seriously!), and they followed up with Deep Research in Gemini that's reimagining how we interact with information. Meanwhile, ChatGPT just kicked off 2025 by finally adding reminder features (though I have some thoughts about that).

Speaking of recent developments, I had the opportunity to speak at the AI Health Trust webinar last week about something that's been on my mind lately - how we're building AI for healthcare and why having the right expertise matters more than perfect metrics. It's a topic that deserves more attention, especially as we're seeing record investments in healthcare AI.

Let's unpack all of this - from quantum breakthroughs to practical AI implementations - and what it actually means for our future.

ChatGPT Joins the Digital Assistant Game - But Is It Really a Game-Changer?

OpenAI just announced a new feature for ChatGPT that lets it function more like a traditional digital assistant - think Alexa or Siri, but with ChatGPT's more sophisticated language capabilities behind it. While this might sound exciting at first glance, let me share my perspective on what this really means.

The new Tasks feature, rolling out to paying subscribers, essentially lets you schedule future actions and reminders. Want a daily weather report at 7AM or a reminder about your passport expiration? ChatGPT can now handle that. It's a nice addition to their toolkit, but here's the thing - this isn't exactly breaking new ground. We've had these capabilities in Alexa, Siri, and Google Assistant for years.

What's interesting isn't the feature itself, but what it signals about OpenAI's strategy. By putting this behind their paywall ($20 and $200 subscription tiers), they're clearly looking to justify their subscription costs. There's also talk about more ambitious developments on the horizon, including "Operator" - an AI agent that could independently control computers, and "Caterpillar" for more complex task management.

But let's be real - while it's a welcome addition for ChatGPT users, this is more about catching up to existing digital assistants than revolutionary innovation. The real test will be whether they can deliver reliable results consistently, something that has been a challenge with their previous agent demos.

For now, if you're a paying ChatGPT user, you can try it out by selecting "4o with scheduled tasks" in your model picker. Just remember it's still in beta - so maybe don't trust it with your most critical reminders just yet.

Source: The Verge 

Google's Quantum Leap Claims to Prove Multiple Universes - Here's Why It Matters

Let me share something that sounds like science fiction but could very well be one of the biggest tech breakthroughs of recent years. Google has unveiled Willow, their latest quantum computing chip, and the implications are mind-bending.

Here's what makes this announcement particularly fascinating: Google's Quantum AI founder, Hartmut Neven, claims the chip is so incredibly fast that it must be borrowing computational power from parallel universes. We're talking about calculations that would take today's fastest supercomputers 10 septillion years (that's a 1 followed by 25 zeros) completing in under five minutes.

But let's keep our feet on the ground for a moment. While some experts are backing these extraordinary claims, skeptics point out that Google is using their own benchmark to measure this performance. It's like creating your own measuring stick - impressive, but perhaps not definitive proof of multiple universes.

What's undeniably significant is the progress in error reduction - a critical challenge in quantum computing. Traditional computers work with simple on/off states (0s and 1s), but quantum computers use qubits that can exist in multiple states simultaneously. The more qubits you add, the more errors typically occur. Google claims Willow has cracked this problem.

Whether or not you believe in parallel universes, this development signals a potential turning point in computing power. If Google's claims hold up, we're looking at computational capabilities that could revolutionize everything from drug discovery to climate modeling.

What's most exciting is that we're witnessing technology push the boundaries of our understanding of reality itself. As a tech leader, I can't help but wonder: what other fundamental assumptions about our world might quantum computing challenge next?

Source: The Crunch 

Google Launches Deep Research in Gemini: AI That Actually Does Your Research

Speaking of Google, they're taking AI assistance in a fascinating new direction with Deep Research, just launched as part of Gemini Advanced. While their competitors focus on making chatbots more conversational, Google is leveraging its search expertise to solve a real problem: the time-consuming nature of deep research.

What makes Deep Research stand out isn't just its ability to search - it's how it approaches research like a human would. Before diving in, it creates a research plan you can review and modify. Then it starts exploring the web iteratively, using what it learns from each search to inform the next one. Think of it as having a tireless research assistant who can analyze dozens of sources simultaneously and compile findings into a comprehensive report.

The real innovation here is in the execution. Instead of just returning search results or generic responses, Deep Research builds understanding over time through multiple search iterations. This means you get a well-organized report with proper citations and sources - something that would typically take hours to compile manually.

Google's timing is particularly interesting. While companies like Anthropic and OpenAI are advancing their models' reasoning capabilities, Google is focusing on making AI practically useful for real-world tasks. For Gemini Advanced subscribers, this means transforming hours of research into a matter of minutes.

This release, combined with their new experimental Gemini 2.0 Flash model, signals Google's strategy to differentiate itself in the AI space by focusing on concrete utility rather than just raw capabilities. It's a glimpse of how AI tools might actually become indispensable research assistants rather than just sophisticated chat partners.

Source: Google

AI in Healthcare 2025: Beyond the Hype, Into Reality

Let me share an interesting perspective on the current state of AI in healthcare, based on PitchBook's latest market snapshot. The numbers tell quite a story, but there's more beneath the surface.

The sector has seen remarkable activity, with $10.5 billion invested across 511 deals in 2024, showing strong investor confidence despite market normalization from the 2021 peak of $22 billion. But here's what makes this fascinating: it's not just about throwing money at shiny new AI tools - we're seeing a fundamental shift in how healthcare operates.

Biotech Takes the Crown 

Biotech is leading the charge, attracting $4.9 billion in recent funding for good reason. Market leaders like insitro and Generate:Biomedicines aren't just raising impressive funding rounds; they're showing real progress in bringing AI-discovered drugs to clinical trials.

Medtech's Quiet Revolution 

The medtech sector, with $2.2 billion in investments, is silently transforming patient care in profound ways. Companies like Tempus AI are revolutionizing disease detection and treatment planning, making early diagnosis more accurate and accessible. What's exciting isn't just the technology - it's seeing these tools actually make their way into everyday clinical settings, from remote patient monitoring to AI-enhanced surgical devices.

Healthtech's Integration Challenge 

Healthtech has secured $3.3 billion, but the story here isn't just about the money - it's about making healthcare smarter and more accessible. From streamlining hospital workflows to personalizing mental health support, these solutions are tackling real-world problems. Companies are focusing on practical applications that can immediately impact patient care and provider efficiency.

Looking Ahead 

What's particularly interesting is how the market has matured. We're seeing investors become more selective, focusing on companies that can demonstrate real clinical value rather than just technological prowess. This shift from pure innovation to practical implementation is exactly what healthcare needs.

The challenge ahead isn't about developing more AI tools - it's about making them work seamlessly within our complex healthcare systems. With sustained investor interest and a growing focus on real-world applications, we're watching AI transform from a buzzword into a fundamental part of healthcare delivery.

Source: PitchBook

Building Better Healthcare AI: Why Domain Expertise Matters

I recently had the opportunity to speak at the AI Health Trust webinar alongside other AI experts, where we touched on an interesting topic I'd like to expand on here - how to effectively incorporate domain experts into AI healthcare development.

The timing for this discussion is perfect. While LinkedIn is flooded with companies showing off impressive AI metrics and benchmarks, there's a growing realization that technical excellence alone doesn't guarantee success in healthcare settings.

During the webinar, I emphasized that we need a transdisciplinary approach - not just multidisciplinary or interdisciplinary. While having medical experts and developers working in parallel is good, and having them collaborate is better, we need them to truly understand each other's domains to create effective solutions.

• Developers need more than coding skills - they must understand why solutions that work in controlled environments might fail in clinical settings
• Validators ensure our models work in healthcare's messy reality, not just perfect test conditions
• Operational Staff provide crucial insights about practical implementation and real-world usage

This is why I advocate for what I call "knowledge exchange platforms." It's not about creating new software - it's about building frameworks where developers can spend time in emergency rooms, and clinicians can participate in AI development discussions. While it might seem inefficient initially, it prevents those "perfect on paper, useless in practice" solutions we see too often.

The future of healthcare AI depends on bridging technical expertise with clinical wisdom. It's not just about having great engineers or experienced doctors - we need people who truly understand both worlds.

Want to hear more insights from our discussion? Check out Health Ai And Technical Trust Full Webinar Recording:

____

That’s all for this month! From Google’s quantum leaps hinting at parallel universes to OpenAI stepping up their digital assistant game, it’s clear 2025 is off to a groundbreaking start. 

And while I’d love to dive deeper into these stories with you, I’ve decided to pause this newsletter for now. I hope to be back soon with more fascinating updates. 

Until then, stay curious and keep exploring!