— Comments Off on 12 Tech Trends & Weak Signals Shaping 2026

12 Tech Trends & Weak Signals Shaping 2026

If we thought things were changing rapidly at the end of last year, the pace of innovation heading into 2026 is unlike any previous cycle. AI is shifting from tool to collaborator. Energy systems are becoming software-defined. Nations are building self-reliant tech stacks. And early-stage breakthroughs in biology, materials, and flight hint at disruptions still in their infancy.

Using the same structure as we did for our 2025 trends forecast, let’s take a look at the trends that matter most for leaders navigating the next 12–36 months.

We organize trends into three categories based on their maturity and impact timeline:

Game Changers — Trends reshaping entire industries right now. These affect how businesses operate, how value is created, and how competitive advantage is built.

Foundational Breakthroughs — Scientific and engineering advances unlocking new possibilities. These create strategic optionality for the decade ahead.

Weak Signals — Early indicators of transformations that will dominate the late 2020s. These require positioning now, even if mainstream adoption is years away.

Game Changers (Reshaping Business in 2026)

1. The Context Shift: AI moves from prompts to persistent understanding

AI is moving beyond “type a prompt, get an answer.” We’re entering an era where systems maintain memory, understand situational context, and adapt to the user — almost like a knowledgeable collaborator who already knows your goals.

Models now handle million-token context windows—enough to reason across entire codebases or product histories in a single session.

What this looks like in practice:

  • Microsoft’s Recall feature reconstructs your past digital activity to support proactive workflows.
  • OpenAI’s ChatGPT now maintains both explicit memories and learns from chat history to deliver personalized responses across sessions.
  • Personal AI applications like Lindy automate entire workflows—triaging inboxes, booking meetings, and updating CRM entries based on your preferences.
  • Otter.ai transforms meeting notes by remembering context from previous discussions to provide more relevant summaries.
  • Meta Ray-Ban Display uses AI, cameras, and the Neural Band to provide information, translations, and navigation relevant to the user’s immediate surroundings, including anticipating needs, suggesting replies, and offering reminders.

Why it matters: Organizations must redesign workflows around AI that already knows your context, not AI that waits for commands.

2. The Agency Economy: Work shifts from doing to directing

One of the biggest shifts heading into 2026 is how human roles are being redefined. For decades, technology focused on productivity—helping us work faster or automate small tasks. Now, as AI agents take on full execution of workflows, human value is moving upstream to direction, judgment, creativity, and decision-making.

A recent Deloitte study found that nearly half of enterprises now use autonomous AI agents in operations. Teams are spending less time performing tasks and more time setting goals, defining constraints, and shaping outcomes. The shift isn’t about humans doing less—it’s about doing fundamentally different things, that only humans excel at.

Why it matters: In the Agency Economy, human value becomes agency itself: the ability to define what work should accomplish, not just how to accomplish it. Organizations that redesign roles around this shift will unlock entirely new levels of impact.

3. The Networked Mind: Hybrid AI spans edge and cloud

AI is evolving into coordinated networks of specialized models working across edge devices and cloud infrastructure. Small, fast models run locally on AI PCs for privacy and speed, while large models in the cloud tackle complex reasoning. Research shows model routing architectures could cut inference costs up to 85% while improving accuracy.

Real-world momentum:

  • HP AI Companion balances local SLMs for privacy with cloud models for complex reasoning.
  • Microsoft Foundry / Azure OpenAI  intelligently routes prompts across OpenAI models, Claude, and other models to optimize cost and performance.
  • Arcee Conductor enable seamless model switching across providers without changing code.
  • Google Gemini Nano is designed to run on mobile devices and leverage larger cloud-based Gemini models for more complex needs.

Why it matters: AI becomes a continuously operating intelligent network rather than a tool you call on demand.

4. Stable Value Rails: Digital dollars become default global settlement

Stablecoins—digital currencies pegged to traditional assets like the US dollar—are becoming core financial infrastructure. In 2024, total stablecoin transfer reached $27.6 trillion, surpassing the combined transaction volume of  Visa and Mastercard by more than 7%.

What’s driving this explosive growth? Faster settlement times (transactions in seconds vs. days), improved liquidity management, and competitive pressure are pushing businesses to adopt stablecoin payment rails. Major financial institutions are recognizing that stablecoins offer 24/7 settlement without geographic constraints.

Recent momentum

  • Circle’s June 2025 IPO saw shares triple on opening day, valuing the USDC stablecoin issuer at $18+ billion and signaling massive institutional confidence.
  • PayPal’s PYUSD and  Stripe’s stablecoin payment integrations bring digital dollars to mainstream commerce.
  • Highnote, an embedded finance platform, enables businesses to integrate stablecoin payments into their apps seamlessly.

Why it matters: By late 2026, expect global commerce to increasingly settle on-chain, whether consumers notice or not.

5. Sovereign Stacks: Nations build independent AI and chip ecosystems

Countries are constructing independent technology stacks—from semiconductor fabs to AI training infrastructure. Since 2021, government incentives in the U.S. and EU have catalyzed over $400 billion in announced semiconductor investments globally, with major production facilities breaking ground in Arizona, Ohio, and across Europe.

Recent movements:

  • China’s DeepSeek models and Harmony OS power domestic alternatives to Western tech.
  • India’s BharatGPT project builds national AI models and cloud infrastructure.
  • U.S. National AI Safety Institutes establish sovereign oversight and capability.

Why it matters: Technology = geopolitics = economic leverage. Global tech companies must navigate increasingly nationalized infrastructure.

Foundational Breakthroughs (Unlocking the Next Decade)

6. Living AI: Models that learn as they go

Today’s AI models know only what they are trained on. In the future AI will be able to continuously learn and adapt in real time — potentially unlocking AI superintelligence. 

Where this is emerging:

  • HP Wolf Security uses adaptive AI to learn from endpoint behavior across millions of devices.
  • Safe Superintelligence reflects a new class of AI companies focused on building systems that can learn and improve over time—without sacrificing safety, alignment, or human oversight.
  • Hazy Research (Stanford) – Is working on adaptive data-centric AI systems that evolve with changing inputs.

Why it matters: Continual learning is the missing piece for AI superintelligence. Today’s models are frozen in time after training. Models that learn continuously can discover new insights, adapt to changing environments, and potentially develop reasoning capabilities beyond their initial programming—transforming AI from a tool that recalls information to one that can genuinely innovate and invent.

7. Omnimodal Intelligence: AI that perceives the entire world

Multimodal AI (text + image + audio) was only the beginning. Omnimodal systems can combine vision, language, spatial data, code, simulation, physics, and robotic action— enabling AI to understand not just our digital worlds but the physical world around us.

In 2025, Google DeepMind’s advanced Gemini with Deep Think achieved gold-medal performance at the International Mathematical Olympiad, solving five out of six problems and earning 35 points . This multimodal breakthrough demonstrated AI’s ability to reason across complex domains—processing mathematical language, geometric representations, and abstract symbolic relationships simultaneously—all within the competition’s 4.5-hour time limit.

Where we’re seeing it emerge:

  • OpenAI Sora2 links language, vision, motion, and sound into a single generative system.
  • WorldLabs Marble generates 3D worlds from text, images, video, and layouts in a frontier multimodal world model.
  • OmniVinci (NVIDIA Research) builds shared vision, audio, and text understanding in a unified omni-modal embedding space.

Why it matters: This is the foundation for robotics, AR, autonomous systems, and digital environments that understand us as richly as we understand them.

8. Fusion Breakthroughs: Commercial fusion enters the transition zone

Fusion is crossing from physics research into engineering reality. This year Helion Energy started construction on the first nuclear fusion powerplant and establishing manufacturing operations to assemble future facilities.

Breakthrough momentum:

Why it matters: Fusion solves AI’s energy bottleneck. Unlimited clean energy would remove the primary constraint on computing advancement, revolutionize manufacturing, and enable breakthroughs from desalination to space exploration. Strategic leaders should model scenarios where baseload energy costs drop 70–80%.

9. Skydriven Mobility: Cities prepare for air-first commuting

Electric air taxis and autonomous aerial systems are moving from prototype to certification. Joby Aviation recently added three additional vertiports to its Dubai’s electric air taxi network.

How it’s taking off:

Why it matters: This mirrors early rideshare—niche at first, then indispensable for urban mobility.

Weak Signals (Position Now for 2028–2030)

10. Responsive Reality: work and spaces assembled around intent

Work is no longer confined to fixed offices, static software, or predefined workflows. As AI becomes more context-aware, it can assemble tools, interfaces, and even physical environments on demand, reshaping workspaces around human intent rather than forcing people to adapt to systems.

Early signals are already visible. In 2024, AI-generated software development environments cut setup time by up to 90%. That same capability is now extending to collaboration, meetings, and hybrid work environments.

In a Responsive Reality, offices become just-in-time spaces — digital and physical — appearing when needed, adapting as work changes, and disappearing when they’re not. The most powerful workplace isn’t a location or a toolset, but a moment of focused intent, dynamically assembled around the people doing the work.

Where we’re seeing it emerge:

  • AI workspaces (e.g., Radiant) — adaptive digital environments that connect context and action.
  • Microsoft Places — intelligent hybrid work planning that allocates real space based on team needs.
  • Superblocks — enterprise AI app generation for on-demand internal tools.
  • AI app builders (Cursor/Lovable/ToolJet/UI Bakery) — platforms enabling tailored software creation.

Why it matters: Organizations that design for shapeshifting software and spaces will move faster, waste less time and real estate, and unlock higher-leverage human work by removing friction between intent and execution.

11. The Elastic Grid: Energy as a dynamic, software-defined network

Energy is shifting from a centralized utility to a decentralized, flexible grid coordinated by AI.

Virtual Power Plants expanded 33% year-over-year in 2024, powered largely by EVs and home batteries contributing energy back to the grid.

Examples in motion:

Why it matters: Energy becomes elastic—expanding and contracting in real-time based on software signals, not just physical infrastructure.

12. The Longevity Stack: health span as a technology platform

Longevity is shifting from supplements to integrated technology stacks—genomics, wearables, metabolic sensors, and AI diagnostics working together to not just make us healthier but live longer than we can imagine today.

In 2025, the longevity and anti-ageing market was valued around USD 85 billion and is forecast to exceed USD 120 billion by 2030. A strong sign that longevity is transitioning from niche science to mainstream investment.

Early examples:

  • Altos Labs and Calico advancing cellular rejuvenation research.
  • Apple Watch and Whoop embedding early-detection biomarker analysis.
  • Biotech–AI research and startups using foundation models to predict disease years in advance.
  • Viome’s RNA-level insight into thousands of biomarkers and using AI to decode your body and help develop a longevity game plan.

Why it matters: As longevity becomes a programmable technology stack, healthcare shifts from reactive treatment to proactive optimization — extending healthy, productive years while reshaping healthcare costs, workforce dynamics, and entire wellness industries.

Looking Ahead

As we look forward,  AI becomes contextual, energy becomes flexible, mobility takes to the skies, and biology turns programmable. These aren’t separate breakthroughs — they’re converging into a new operating system for society, redefining how work happens, how economies function, and how humans extend their capabilities. Organizations won’t compete on access to technology alone, but on how well they design for convergence — connecting intelligence, infrastructure, and human agency into systems that can adapt continuously.

The leaders who thrive in this next chapter won’t simply adopt new tools. They’ll design for what only humans bring: judgment, creativity, empathy, values, and the ability to imagine futures that don’t yet exist. As AI and machines accelerate execution, our humanness becomes the scarce advantage — the compass that guides intelligent systems toward outcomes worth pursuing.

The future isn’t something that happens to us.
It’s something we choose to shape.
In a world where machines learn fast, our advantage isn’t speed — it’s soul.

Blog Futurism & Technology Trends
— Comments Off on How corporate venture capital de-risks emerging technology development

How corporate venture capital de-risks emerging technology development

As the pace of technological change accelerates, corporations are no longer just adapting to disruption — they are actively investing in it. Through their corporate venture capital (CVC) arms, established companies are entering partner-investor relationships with startups across frontier domains such as artificial general intelligence (AGI), humanoid robotics, quantum computing, and other potentially game-changing technologies.

In this post, I’m exploring how corporates are using CVC to de-risk exploration in these emerging tech domains, the competitive advantage of corporate-startup partnerships, which categories are particularly well suited to CVC versus traditional VC, and how emerging-tech startups can best position themselves for CVC investment.

How CVC reduces technology risk for corporations

Corporations face an inherent tension: they need to pursue innovation and new business models, but they also must manage risk, protect core business margins, and maintain operational stability.

That’s why more than 25% of the funding deals last year included CVCs. CVC offers a hybrid path: by investing in and partnering with startups in nascent and emerging technologies, corporates can explore adjacent or disruptive opportunities while externalizing much of the technology risk.

In domains such as AGI, quantum computing, space tech, or next-gen energy storage, the technologies are capital-intensive, have long horizons, involve high technical and commercialization risk, and often require domain-specific assets, manufacturing, regulatory engagement, or ecosystem partnerships. For corporations, investing via CVC is a strategic way to gain early exposure, build optionality, secure technology rights or vantage, and integrate promising startups into their ecosystem — enabling them to stay ahead of both disruptive threats and complementary opportunities.

However, not all companies should invest in all emerging technologies — the key is strategic selectivity, focusing on technologies that could either disrupt their industry or offer complementary capabilities to enhance their competitive position.

CVC is a de-risking engine for corporates exploring emerging tech — it lets them access options in high-uncertainty spaces without the full burden of building in-house, while building strategic alignment with their core business and future growth vectors.

How CVCs reduce technology risk through strategic value creation

While most emerging tech startups will need venture capital funding, nearly all funding deals can benefit significantly from CVC participation. The question isn’t whether corporate venture capital adds value, but rather how CVCs uniquely reduce technology risk for both corporates and startups across different technology categories.

What makes CVC different from traditional VC?

CVC brings strategic value beyond capital: manufacturing capabilities, distribution networks, supply chain access, regulatory expertise, and direct integration pathways. Traditional VC focuses primarily on financial returns and rapid scaling without operational entanglement.

Especially suited for CVC

Best for: Hardware, long horizons, strategic fit, high-capex technologies

  • Hardware + embedded systems (e.g., humanoid robots, advanced compute, quantum computing, next-gen energy storage, nuclear energy) — These domains require supply chain, manufacturing, and integration with existing platforms, often with regulatory or domain-specific partnerships. Corporations with manufacturing or platform assets can add real value. For example, HP Tech Ventures’ investment in EdgeRunner AI demonstrates how corporates can accelerate AI hardware integration. EdgeRunner builds domain-specific, air-gapped, on-device AI agents for military and enterprise applications that operate entirely without internet connectivity. The company’s platform delivers mission-specific AI assistants that ensure low latency, enhanced data privacy, and reduced cloud costs — critical advantages that scale when coupled with AI hardware platforms and edge computing products and expertise. Similarly, Intel Capital’s investment in Rigetti Computing showcases how corporate backing accelerates quantum computing development. Rigetti builds full-stack quantum computers, and Intel’s expertise in chip manufacturing, supply chain access, and deep semiconductor knowledge provides strategic advantages that pure financial investors cannot match, reducing both technical and commercialization risk.
  • Platform or ecosystem technologies (technologies that require broad industry adoption and create value through network effects, such as 6G/hyperconnectivity, clean-tech infrastructure etc.) — Corporates are often deploying or will deploy these platforms themselves, so investing via CVC gives them inside access and optionality.
  • Strategic technology adjacencies for the corporate. For example, if a corporate sees synthetic biology or biotech as a future adjacency to their business, then CVC allows them to explore while leveraging internal capabilities (e.g., R&D, supply chain, global operations).
  • High-capex / long-horizon technologies — Traditional VCs demand high returns within a fixed timeline, but corporates can afford longer horizons if strategic alignment is strong.

Related to the above, trending data show that CVCs have recently been prioritizing AI and Robotics, which exemplify both platform technologies and strategic adjacencies that many corporates are exploring. For example, nearly 30% of CVC deals in 2024 revolved around AI.

HP Tech Ventures’ recent investment in Multiverse Computing — a quantum-inspired AI company that compresses large language models by up to 95% while maintaining performance — exemplifies this trend. Multiverse’s technology addresses a critical infrastructure challenge in AI deployment, enabling models to run on edge devices and dramatically reducing computing costs and energy consumption.

HP’s strategic support helps Multiverse scale this technology across enterprise applications, bringing AI benefits to companies of all sizes.

Making it work for both CVC and startup

In the accelerating wave of frontier technologies — from AGI and quantum computing to next-gen energy storage, synthetic biology, and space tech — the smart corporates will not wait passively. They will deploy their CVC as a strategic lever to access, partner with, and accelerate startups that can redefine their future business models.

How do startups benefit from CVC partnerships?

For startups operating in these domains, the path to growth means not just securing capital, but forging the right strategic partnerships: ones that bring scale, integration, and access to a corporate ecosystem that would otherwise take years to build.

For HP Tech Ventures, this means offering portfolio companies access to HP’s world-class technology, one of the world’s largest channel and distribution partner networks, and a vast global manufacturing and supply chain — resources that help startups scale rapidly and achieve significant market impact.

Emerging tech sectors alignment

From humanoid robots to synthetic biology, the next wave of innovation is rewriting the boundaries of what’s possible — and CVCs are uniquely positioned to shape that future. The following table maps how each major emerging-tech sector aligns with corporate venture capital, and what founders should keep in mind as they navigate this evolving landscape.

What should founders consider when pursuing CVC?

By aligning technology, business model, partner strategy, and timing, both corporations and startups can ride the emerging-tech wave with lower risk and higher impact.

If you’re a startup in one of these frontier domains and are thinking about CVC, ask yourself: Which corporations in my value chain have scale, distribution, or manufacturing that could accelerate me? How much risk are they willing to take? Am I ready for strategic integration?

Blog corporate venture capital
— Comments Off on 2025 Weak Signal Wild Cards: Quantum Computing, 6G Networks and Hyperconnectivity, and Space Tech

2025 Weak Signal Wild Cards: Quantum Computing, 6G Networks and Hyperconnectivity, and Space Tech

At the beginning of the year, I outlined 10 technology trends and weak signals I felt would have a transformative impact on 2025 and beyond. These emerging innovations represent not just incremental improvements but potential paradigm shifts that could fundamentally alter industries, economies, and societies.

These trends fall into three categories for me:

  1. Game Changers are set to have a significant impact on industries, societies, and markets in 2025 and beyond. Will transform how we work, learn, and live.
  2. Foundational Breakthroughs are major technological advancements needed for game changer technologies to succeed.
  3. Weak Signal Wild Cards present the opportunity to be a future game changer or a foundational breakthrough but still in a nascent stage with a number of headwinds to overcome.

Today, I’m diving into Weak Signal Wild Cards — Quantum Computing, 6G Networks and Hyperconnectivity, and Space Tech, and the opportunities and the questions they raise.

Quantum Computing Maturation: The Next Computing Revolution

Quantum computing is progressing toward achieving quantum advantage, the ability to solve complex problems that are beyond the reach of classical computers.

The Promise of Quantum Applications

The potential applications are transformative. In drug development and materials science, quantum computers can simulate chemical reactions with unprecedented efficiency, potentially accelerating pharmaceutical innovation by years. This capability stems from quantum computers potential ability to model molecular behavior at scales that would take classical computers millennia to process.

Major breakthroughs in error correction, qubit stability, and chip miniaturization are rapidly advancing commercialization efforts. Companies like IBM and Google predict that 1,000+ logical qubit systems will be operational by 2030, a milestone that could unlock entirely new categories of problems to solve.

However, the threat of quantum computers breaking asymmetric cryptography — the algorithms that our digital world relies on — grows every year. Experts think there’s up to a 34% chance of this happening by 2034. This would put encrypted communications at risk, compromise the existing digital signatures used to verify the integrity of firmware and software, and undermine digital trust.

HP recently launched the world’s first printers to protect against future quantum computer attacks. Without quantum resilience, a printer facing a quantum attack at the firmware level would be fully exposed through malicious firmware updates, allowing the attacker to achieve stealthy, persistent, and total control of the device.

Weak Signals to Watch

Several weak signals suggest where quantum computing might head next. Breakthroughs in error correction and scalability are occurring alongside a push for quantum cloud services, which could democratize access to quantum computing power. More intriguingly, cross-disciplinary collaborations in material science might lead to unforeseen quantum applications beyond current predictions.

The Unexpected Outcome

Early breakthroughs in cryptography or molecular simulations could transform industries like cybersecurity and drug development far sooner than anticipated. The cryptographic implications alone are staggering — quantum computers could potentially break current encryption standards, necessitating a complete overhaul of digital security infrastructure. This creates both a threat and an opportunity, driving the development of quantum-resistant cryptography.

The question isn’t whether quantum computing will change the world, but rather: Are we prepared for how quickly it might happen?

6G Networks and Hyperconnectivity: Beyond the Speed Barrier

While 5G networks are still rolling out globally, researchers and telecommunications companies are already laying the groundwork for 6G, and the implications extend far beyond faster download speeds.

The Speed Revolution

6G recently clocked speeds 10 times faster than 5G. This speed enhancement will fundamentally enhance the Internet of Things (IoT), real-time augmented and virtual reality, and autonomous systems. Commercial 6G rollouts are expected in the 2030s, though pilot programs and test networks are already being established.

To put this in perspective, 6G could enable instantaneous data transfer at rates approaching 1 terabit per second. This is the difference between streaming a movie and downloading an entire library in milliseconds.

Weak Signals on the Horizon

Integration of 6G with AI-driven networks promises intelligent, self-optimizing communication systems that can predict and respond to network demands in real time. Energy-efficient hardware development is critical, given the massive power requirements of hyperconnectivity.

The Unexpected Outcome

Perhaps the most compelling possibility is the emergence of real-time brain-computer interfaces (BCIs) leveraging 6G’s ultra-low latency and massive bandwidth. Direct neural connections to digital systems could revolutionize how we interact with technology, enabling thought-controlled devices, instant information retrieval, and even neural-to-neural communication.

This raises profound questions: What happens when the boundary between human cognition and digital networks blurs? How do we ensure equitable access to such transformative technology?

Space Tech and Industry Expansion: The Final Frontier Goes Commercial

Space exploration is transitioning from government-led missions to commercial dominance, with private companies pushing boundaries in space tourism, resource extraction, and even permanent settlements.

The Economics of Space

By 2030, the space economy could exceed $1 trillion, driven by dramatically lower launch costs and the prospect of resource extraction from asteroids and other celestial bodies. SpaceX’s reusable rocket technology has already reduced launch costs by an order of magnitude, opening space to a new generation of commercial ventures.

This economic shift is enabling applications that were science fiction a decade ago: space-based manufacturing in zero gravity, asteroid mining for rare earth elements, satellite mega-constellations for global internet coverage, and the early stages of space tourism. Not to mention the possibility of AI data centers in space powered by the sun and connected to the earth via near-instantaneous 6G network connectivity!

The Key Players

SpaceX, Blue Origin, and Rocket Lab lead the private-sector charge, while NASA, the European Space Agency, and China’s National Space Administration (CNSA) continue to push the boundaries of science and exploration. The interplay between commercial innovation and government-funded research is creating a unique ecosystem where public and private interests simultaneously align and compete.

Weak Signals in Orbit

Development of in-orbit manufacturing capabilities could enable the construction of structures too large or complex to launch from Earth. Asteroid mining technologies are advancing from theoretical to practical, with several companies working on prospecting missions. These capabilities could fundamentally alter Earth’s resource economics and manufacturing paradigms.

The Unexpected Outcome

The rise of space militarization or territorial disputes over space resources presents a darker possibility. As space becomes economically valuable, the potential for conflict increases. Who owns asteroid resources? What are the rules of engagement in orbit? How do we prevent an arms race in space?

The Outer Space Treaty of 1967 was written for a different era. As commercial interests expand beyond Earth, we may need new frameworks for space governance, resource allocation, and conflict resolution.

The Convergence Question

These three wild cards don’t exist in isolation. Quantum computing could enable the complex calculations required for 6G network optimization and space mission planning. 6G networks could provide the bandwidth necessary for controlling quantum computers remotely or coordinating space operations in real time. Space-based quantum communication networks could create unhackable global communication systems.

The real transformation may come from the technologies’ convergence.

Looking Forward

Weak signal wild cards are inherently uncertain. They represent technologies with transformative potential but significant obstacles to overcome, including technical challenges, regulatory hurdles, public acceptance, and economic viability. These all factor into whether these nascent technologies become foundational breakthroughs or game-changing forces.

What makes this year particularly interesting is that we’re at an inflection point for all three. Quantum computers are moving from laboratories to commercial applications. 6G standards are being defined and tested. Space ventures are scaling from experimental to operational.

The next few years will determine whether these weak signals amplify into game-changing technologies or encounter obstacles that delay their impact. Either way, they’re worth watching closely because when wild cards come in, they rarely announce themselves in advance.

Blog Futurism & Technology Trends