By Werner Vogels, CTO at Amazon
For much of the world, technology has become so intertwined with our day-to-day lives that it influences everything. Our relationships, the care we seek, how we work, what we do to protect ourselves, even the things we choose to learn and when. In the coming year, we will begin the transition into a new era of AI in the human loop, not the other way around. This cycle will create massive opportunities to solve problems that truly matter. And it starts by addressing one of the unintended consequences of our hyperconnected world—loneliness and a lack of companionship—by turning the very force that created the problem into the solution.
Companionship is redefined for those who need it most
Loneliness has reached epidemic proportions, affecting 1 in 6 people worldwide and designated as a public health crisis by the World Health Organization. We stand at the threshold of a profound transformation in human-technology relationships, one that directly addresses this loneliness epidemic through genuine emotional connection.
The convergence of aging demographics, advanced AI capabilities, and a global loneliness epidemic have created the perfect conditions for a companionship revolution. We are witnessing a shift from transactional device interactions to relationship-building with physical AI that demonstrates increasingly nuanced emotional intelligence and responsive behaviors. Clinical evidence supporting the effectiveness of combating loneliness with companion robots is compelling.
We’re biologically hardwired to project intent and life onto any movement in our physical space that seems autonomous to us. This biological response creates the foundation for companion robots to provide the consistent emotional presence that alleviates loneliness.
Rather than replacing human caregivers, this companion revolution creates a collaborative model where technology and people work in tandem to provide care and fight loneliness. Robots will handle routine monitoring and provide steady emotional presence, while allowing humans to focus on complex decision-making and nurturing deeper relationships. As people form deep trust with these robotic companions, the companies building them must implement strong controls to ensure these robots never exploit that trust to influence users’ decisions or shape their beliefs. When developed responsibly with these safeguards in place, this represents technology at its best: keeping people central to care while extending our capacity to support those who need it most.
The dawn of the renaissance developer
Tools change, but the fundamentals endure. As GenAI reshapes how we build software, a familiar trope has re-emerged, the narrative that developers will become obsolete. This is not the end of the developer, it’s the dawn of something new, the renaissance developer.
Time and time again we have seen that lowering the barrier for entry doesn’t eliminate the need for human expertise, it amplifies it. GenAI lets us generate code in seconds, but if you put garbage in, you get really convincing garbage out. The politics, the constraints, the unspoken priorities that shape every technical decision are nuanced and require a developer who understands why it matters to the humans who pay for it and the humans that will use it.
Developers can no longer live in silos. You must think bigger, the moment demands it. This is the dawn of a new age for developers. You have never been more valuable. Your creativity has never been needed more. So keep building, stay curious, and keep solving the world’s hardest problems.
Quantum-safe becomes the only safe
Personal data, financial records, and state secrets are already being harvested by malicious actors betting on quantum’s arrival. For most organizations, the reasonable assumption was that they had years to plan. That assumption no longer holds. Advances in error correction and algorithmic efficiency have compressed, and the window for proactive defense is closing. The coming year requires post-quantum thinking; from cryptography protecting our most sensitive communications to the education necessary to train quantum engineers.
The risk lies in how we secure data today. Malicious actors have been harvesting encrypted data for years, patiently waiting for the computing power necessary to decrypt it. Most of our digital security relies on public-key cryptography, and the mathematical puzzles that make RSA and elliptic curve encryption hard for classical computers to solve will be trivial for quantum machines running algorithms like Shor’s. Unlike symmetric encryption, which can be strengthened with longer keys, public-key systems need entirely new mathematical foundations to survive the quantum era.
Preparation isn’t something you can put off, the work must begin now, and organizations need to act on three fronts: deploying post-quantum cryptography (PQC) where we can, planning to update and replace physical infrastructure where we can’t, and developing quantum ready talent to support this transition.
The good news is that PQC solutions exist and are deployable now at the OS level, the browser level, and in the cloud. Major tech companies are converging on NIST standards, ensuring interoperability and security. Detailed migration plans exist.
The physical world is where the transition becomes most complex. Millions of smart meters use current encryption standards but lack the processing power to run post-quantum algorithms. Power grids, water treatment systems, and transportation networks face similar constraints with embedded devices that cannot be easily upgraded. This constraint will force companies to get creative. Expect hybrid approaches that layer quantum-safe gateways in front of legacy devices. This is a cross-functional transformation.
Finally, there’s talent. Organizations that invest in quantum education and training now will build competitive advantages that cannot be easily replicated. The quantum era requires a new blend of expertise that is rare today, but will be table stakes in the next few years.
Defense technology changes the world
Military investment in technology is surging, both by governments and in the private sector. The speed of innovation has significantly increased, and in the coming years we will see the timeline from battlefield to civilian application compressed.
Companies like Anduril Industries and Shield AI design technologies as dual-use from inception, seeing civilian applications not as afterthoughts but as core business models. This shift eliminates the traditional adaptation phase that historically added years to the transfer timeline. This creates feedback loops measured in days rather than decades.
Beyond conflict zones, night vision systems now guide search-and-rescue. Tactical edge computing powers remote healthcare clinics. Autonomous systems developed for military logistics are adapted to solve agricultural labor challenges, with immediate applications in power plants, wind farms, search and rescue operations, and maritime port security.
Healthcare systems, emergency services, and infrastructure operators should prepare for capabilities that will emerge from current defense investments within the next two years, not two decades. The organizations that understand this accelerated timeline will gain significant advantages in solving critical problems.
The technologies being refined under extreme pressure today won’t wait for peacetime to reach the masses. They’re arriving now, designed from the start to serve both military and civilian needs..
Personalized learning meets infinite curiosity
AI has the power to fundamentally change the way that we approach education. Instead of forcing every student through the same system, AI will adapt to how each child thinks.
A student can now access tutoring from an AI system for $4 per month. Khan Academy’s Khanmigo exceeded all projections by 1,400%, reaching 1.4 million students in its first year. Anthropic launched the world’s first nation-wide AI education pilots in Iceland. These aren’t experiments—they’re production systems at scale. This transformation is happening in India, Brazil, and across Africa.
To be clear, teachers are NOT going away. Teachers who use AI tools save an average of 5.9 hours per week, which equates to about six weeks per school year. It’s also allowing educators to reach more students even with tight financial constraints.
In 2026 and beyond, personalized AI tutoring will be as ubiquitous as smartphones. When you use tools to engage curiosity instead of enforcing compliance, when you honor diversity instead of demanding conformity, schools spring to life. And that changes everything.
