DD
MM
YYYY

PAGES

DD
MM
YYYY

spot_img

PAGES

Home Blog Page 87

Samsung Unveils First Agentic AI Phone

0

Samsung has officially unveiled the Galaxy S26 series including the S26, S26+, and S26 Ultra at its Galaxy Unpacked event in February 2026, positioning it as the first “agentic AI phone”.

This means the device goes beyond passive AI responses; like answering questions to actively perform tasks on the user’s behalf—handling multi-step workflows, interacting with apps, and executing actions autonomously.

The phone features a multi-agent AI ecosystem under Galaxy AI, integrating three main AI systems: An upgraded, more conversational Bixby; Samsung’s in-house assistant, now powered in part by Perplexity’s APIs for search and reasoning.

Google’s Gemini for advanced agentic capabilities, such as autonomously operating third-party apps like booking rides via Uber. Perplexity as a dedicated third-party AI agent, marking a significant partnership. Perplexity’s integration is particularly deep and system-level—the first time a non-Google company has received such OS-level access on a Samsung device.

A custom wake phrase: “Hey Plex” for hands-free activation. Quick access via side button press-and-hold or other controls. Direct interaction with native Samsung apps like Notes, Calendar, Gallery, Clock, and Reminders; read and write capabilities for seamless multi-step tasks.

Support for select third-party apps. Powering real-time, grounded web search and reasoning across assistants including backend support for Bixby. This setup allows users to choose or switch between agents for different needs, emphasizing flexibility, reduced manual effort, and proactive intelligence.

Samsung describes it as the “beginning of truly agentic AI,” with the phone handling complex background tasks so users focus on results. Perplexity itself highlighted the collaboration as embedding its APIs deeply into the world’s largest Android ecosystem, enabling accurate, orchestrated actions across search, reasoning, and device controls.

The Galaxy S26 series also includes other AI enhancements like improved video editing, noise reduction for calls, and privacy features; a toggleable privacy display on the Ultra model. Pre-orders and expert reviews and unboxings are already circulating following the announcement.

Samsung’s multi-agent approach lets you pick or switch between AI assistants tailored to specific needs, rather than being locked into one like Siri on iPhone or a dominant Gemini on stock Android.

Perplexity stands out for accurate, citation-backed search and reasoning—ideal for research, fact-checking, or complex queries. Its system-level access; first non-Google third-party to get this deep allows it to read/write directly in native apps like Notes, Calendar, Gallery, Clock, and Reminders, enabling seamless multi-step workflows.

Gemini handles true agentic actions, like autonomously navigating third-party apps. Bixby benefits from Perplexity’s backend for better conversational responses and real-time web grounding. This reduces friction in daily tasks: less app-switching, more natural voice commands, and background automation.

Surveys cited by Samsung show 8 in 10 users prefer multiple AI options, so this could make the phone feel like a “trusted companion” that anticipates needs. Privacy-focused users get toggles like the hardware privacy display on the Ultra. Early hands-on impressions suggest Perplexity could be a “sleeper hit” for power users who value grounded answers over creative fluff.

This marks a shift toward an open multi-agent ecosystem on Android, contrasting Apple’s closed “walled garden” Siri approach.Samsung positions Galaxy AI as an “orchestrator” coordinating multiple AIs, reducing reliance on any single provider even as Gemini remains central for many agentic features.

Granting Perplexity OS-level access breaks precedent—previously reserved for first-party tools—signaling Samsung’s push for platform-level innovation that’s inclusive and adaptable. It differentiates Samsung from Google Pixels and could pressure other Android OEMs to follow suit with multi-agent support.

This partnership is a massive validation and exposure boost.Integration into the world’s largest smartphone maker (Samsung) puts Perplexity’s APIs and “answer engine” in hundreds of millions of potential hands—far beyond prior deals like Motorola preloads or niche carriers.

Powering both a dedicated agent (“Hey Plex”) and backend for Bixby elevates Perplexity from niche search tool to core mobile intelligence player. It underscores Perplexity’s strengths in accuracy, orchestration, and non-hallucinating responses, potentially accelerating adoption and partnerships.

The “agentic AI phone” label highlights a pivot from passive chatbots to proactive agents that act autonomously. Expect more multi-agent phones, with possibilities for additional partners. It accelerates the trend of AI as infrastructure—phones become execution platforms where users focus on intent, and agents handle the “how.”

Privacy and security questions arise with deeper app access, though Samsung emphasizes controls and on-device processing where possible. Apple may respond with enhanced Siri agents; Google could expand Gemini’s reach.

The Galaxy S26 doesn’t reinvent the smartphone form factor but redefines its intelligence layer. If agentic features deliver reliably, it could set a new standard for what “smart” means in 2026—turning phones into true personal agents rather than just devices that respond to commands.

For many, Perplexity’s integration might prove the most practical upgrade for everyday productivity and research. Pre-orders are live, with wider availability starting in March—early reviews will soon reveal how these implications play out in real use.

This move signals a shift toward “AI phones” rather than just smartphones, with agentic capabilities that could redefine how users interact with devices daily. If you’re eyeing an upgrade, the Perplexity integration might be a standout “sleeper feature” for research-heavy or search-focused users.

Google Unveils Nano Banana 2, Its Most Advanced Image Generation Model Yet

0

Tech giant Google has announced the launch of Nano Banana 2, a significantly upgraded version of its image generation model designed to deliver faster performance, sharper visuals, and more realistic outputs.

Also referred to as Gemini 3.1 Flash Image, the model is set to become the default image generator across the Gemini app and other Google platforms. The company stated that the launch signals “a new and vibrant era of generative creativity,” describing Nano Banana 2 as a state-of-the-art image generation and editing system that brings professional-grade capabilities at high speed.

Announcing the launch, Google wrote via a blogpost,

“Today, we’re entering a new and vibrant era of generative creativity with Nano Banana 2. Nano Banana 2 is our state-of-the-art image generation and editing model. It delivers Pro-level image generation and editing at the speed you expect from Flash — making the quality, reasoning, and world knowledge you loved about Nano Banana Pro more accessible. To build a creative that stands out, you need models that naturally integrate into your workflows and scale with ease. Whether you’re building the next major consumer app or marketing campaign”.

Nano Banana 2 introduces several major enhancements. The model is powered by real-time information and visual data from web search, enabling more accurate and context-aware image creation for applications such as education, localized marketing, and travel services.

It also unlocks advanced creative features, including improved text rendering, translation support, and high-resolution upscaling up to 2K and 4K. All images produced with the model will include a SynthID watermark, ensuring clear identification of AI-generated content.

In addition, the system offers native support for multiple aspect ratios, including 16:9, 9:16, and 2:1, delivering richer textures, vibrant lighting, and production-ready visual outputs. Google emphasized enterprise-level transparency through the integration of SynthID watermarking with C2PA Content Credentials, allowing users to understand not only whether AI was used but also how it was applied in the creative process.

The model will soon be available on Vertex AI to help enterprise teams generate more accurate visuals. Google also confirmed expanded integration with Adobe Firefly, strengthening its collaboration with Adobe. According to Steve Newcomb, VP of Product, Adobe Firefly, the integration sets a new standard for image quality and creative control, enabling seamless production-ready workflows.

“With Nano Banana 2 in Adobe Firefly, we’re continuing to bring the industry’s leading AI models together in one creative studio alongside Adobe’s best-in-class tools. Nano Banana 2 in Firefly sets a new standard for image quality and precision, giving creators greater control to generate and refine production-ready work in a seamless workflow. We’re excited to expand our partnership with Google to power even more creative flexibility.”  Steve.

Nano Banana 2 is rolling out across AI Mode and Lens in Search, Flow, and preview access via Google AI Studio and Vertex AI. The release builds on the rapid adoption of Nano Banana, which attracted 13 million first-time users to the Gemini app within four days of its debut and generated more than five billion images within weeks.

The upgrade further strengthens Google’s position in the global AI race, intensifying competition with industry rivals such as OpenAI. This competitive dynamic reflects a broader shift toward platform consolidation, where major technology players are racing to establish end-to-end AI ecosystems that span research, deployment, and commercial application.

Vitalik Reveals Ethereum’s Post-Quantum Roadmap

0

Ethereum co-founder Vitalik Buterin has outlined a detailed “quantum resistance roadmap” for Ethereum, addressing the long-term threat posed by quantum computers that could potentially break current cryptographic systems like ECDSA and BLS signatures.

This announcement came via a comprehensive post on X, building on the Ethereum Foundation’s recent efforts, including establishing a dedicated post-quantum research team earlier in the year. Practical quantum computers capable of breaking modern cryptography don’t exist yet, but progress in the field has raised concerns, with some warnings suggesting risks could materialize by around 2028.

The roadmap focuses on proactively upgrading Ethereum’s vulnerable components over the coming years, integrating these changes into broader Layer 1 improvements like faster finality and scalability. Vitalik highlighted four main quantum-vulnerable parts of Ethereum: Consensus-layer BLS signatures; used by validators.

Data availability mechanisms relying on KZG commitments and proofs. EOA signatures (ECDSA, the standard for externally owned accounts and wallets). Application-layer zero-knowledge proofs; those using KZG or Groth16, common in privacy tools, L2s, and apps.

The plan involves step-by-step replacements with quantum-resistant alternatives, primarily hash-based signatures; variants of Winternitz or similar schemes, STARKs for aggregation and proofs, and native account abstraction. Consensus-layer signatures: Replace BLS with hash-based signatures. Use STARKs for efficient aggregation in full “Lean consensus.”

An intermediate “Lean available chain” step could come sooner with fewer signatures per slot. Selecting a final hash function; options include enhanced Poseidon variants, Poseidon1, or BLAKE3 for efficiency and security. Shift from KZG-based erasure coding to STARK-based alternatives.

Stick to maximized 1D DAS rather than pushing for 2D, given Ethereum’s scale goals. Recursive STARKs handle proofs for correct blob construction. EOA signatures: Introduce native account abstraction; building on proposals like EIP-8141 to support any signature scheme.

Current quantum-resistant options (hash-based) verify at higher gas costs ~200k vs. ECDSA’s 3k, but lattice-based schemes could improve with vectorized math precompiles. Long-term: Protocol-level recursive aggregation to minimize overhead.

Current ZK-SNARKs cost 300-500k gas; quantum-resistant STARKs could hit 10M gas. Protocol-layer recursive signature and proof aggregation via “validation frames” in transactions (EIP-8141). These frames can be replaced by compact STARK proofs, potentially verified at the mempool level; one proof every 500ms rather than on-chain per transaction.

This ties into the Ethereum Foundation’s “Strawmap”, which envisions roughly seven hard forks over ~4 years (every ~6 months), with upgrades like Glamsterdam and Hegotá in 2026. Goals include faster block production and finality (seconds), higher throughput, and full post-quantum security alongside privacy enhancements.

The approach emphasizes gradual migration, engineering feasibility, and avoiding disruption, while acknowledging trade-offs like proof sizes and gas costs. It’s part of a broader push toward “Lean Ethereum” — simpler, more secure, and future-proof. This development reinforces Ethereum’s proactive stance on long-term security.

While quantum computers capable of breaking current cryptography remain years away, this proactive stance positions Ethereum as forward-thinking in long-term security. The roadmap targets four vulnerable areas; consensus BLS signatures, KZG-based data availability, ECDSA for EOAs/wallets, and certain ZK proofs like Groth16/KZG.

Replacing them with quantum-resistant alternatives—primarily hash-based signatures, STARKs for proofs and aggregation, and native account abstraction—aims to make Ethereum resilient against future quantum attacks like Shor’s algorithm. This could prevent catastrophic risks, such as stolen funds from exposed private keys or disrupted consensus.

Vitalik noted that even if quantum threats arrive early, the chain would “keep chugging along” with reduced finality guarantees but no halt. Hash-based/STARK solutions increase proof sizes and gas costs initially ~200k gas for signatures vs. ECDSA’s 3k; STARKs potentially 10M gas without aggregation.

Protocol-layer recursive aggregation via EIP-8141 “validation frames” and mempool-level proving could offset this, keeping costs near-zero long-term. The roadmap bundles this with broader goals: block times down to ~2 seconds, finality in 6-16 seconds from ~16 minutes, gigagas/sec L1 throughput, and teragas L2 scaling—making Ethereum faster and more scalable overall.

Proactive quantum hardening signals Ethereum prioritizes durability, encouraging builders in DeFi, L2s, privacy tools, and ZK apps. It integrates with privacy enhancements and formal verification, potentially attracting more institutional-grade development. However, heavier computations could temporarily raise costs for ZK-heavy apps until aggregation matures.

Bridges, oracles, and cross-chain systems remain vulnerable points (as noted in related discussions). Ethereum’s approach could set a standard, pressuring other chains; Bitcoin lacks a similar detailed plan to follow. Projects like Naoris Protocol highlight full-stack quantum resilience as a growing focus.

This bolsters Ethereum’s case as a secure, future-proof settlement layer for trillions in value. It differentiates ETH from competitors and aligns with institutional interest from BlackRock, Goldman. No immediate price surge is evident from recent coverage, but it strengthens the “ultrasound money” and durability thesis.

By addressing existential threats early, it reduces tail risks for holders. Vitalik’s ongoing ETH sales (noted in context) appear unrelated, tied to personal and portfolio management. Multiple forks introduce upgrade risks though Ethereum’s history shows smooth transitions. Gas cost increases during transition could affect user experience temporarily.

This roadmap reinforces Ethereum’s maturity—treating quantum resistance as a core engineering priority rather than a panic fix. It ties into the “Lean Ethereum” vision: simpler, faster, more secure, and privacy-focused. If executed well by ~2029-2030, it could cement Ethereum’s lead in institutional and long-horizon adoption.

Morgan Stanley Confirms Bitcoin Trading, Lending, Yield And Custody Plans

0

In a significant step toward deeper integration of digital assets into traditional finance, Global Investment Bank Morgan Stanley has confirmed plans to offer comprehensive Bitcoin services directly to its clients.

The bank’s Head of Digital Asset, Amy Oldenburg made the announcement during a conversation with Phong Le President and CEO of Strategy, at the recent Bitcoin for Corporations conference held in Las Vegas in late February 2026.

According to Oldenburg, Morgan Stanley intends to build its own in-house technology for Bitcoin custody and trading, moving beyond the current setup where clients can access spot Bitcoin ETFs and, in some cases, limited crypto trading via the bank’s ETrade platform.

“We will definitely do it,” Oldenburg stated when addressing future plans for Bitcoin custody and trading infrastructure. She further emphasized that developing these capabilities natively is a priority, allowing clients to hold legal custody of their Bitcoin under Morgan Stanley’s regulated oversight.

While some Bitcoin holders particularly long-term advocates may prefer self-custody, Oldenburg noted that bringing assets onto the bank’s platform would unlock additional services in a trusted, institutional environment.

The ambitions don’t stop at basic custody and spot trading. When asked about Bitcoin-based yield and lending products,

“That’s part of the discussion and exploration. It’s a natural part of the roadmap to continue to explore. We’re in a very early journey on that, just in terms of the number of products that are out in the market”, Oldenburg expressed strong support.

She highlighted unexpected momentum in decentralized finance (DeFi) lending protocols and related crypto-native yield opportunities this year, signaling that Morgan Stanley is closely tracking these developments as it considers similar offerings for its clients.

Morgan Stanley’s crypto journey has accelerated noticeably.  Earlier phases allowed select high-net-worth clients to access spot Bitcoin ETFs. In 2025–2026, the firm expanded crypto trading access via E Trade for certain users.

The year 2026 brought the appointment of Amy Oldenburg to lead digital asset strategy, along with filings related to Bitcoin and Solana ETFs in some contexts.

Now, the bank is shifting toward full-stack, proprietary infrastructure rather than relying solely on third-party partnerships.

With nearly $9 trillion in assets under management (and trillions more in advisory relationships), Morgan Stanley’s move could expose millions of traditional investors, wealth managers, and institutions to Bitcoin in a familiar, regulated wrapper.

This development follows similar expansions by competitors such as Goldman Sachs, JPMorgan Chase, Citigroup, BNY Mellon, amongst others, that have moved aggressively into crypto infrastructure, intensifying competition across traditional finance.

Goldman Sachs

Goldman Sachs has built a dedicated digital assets division, offering crypto trading for institutional clients and expanding structured products tied to digital assets. The firm has also explored tokenization initiatives and blockchain-based financial instruments.

JPMorgan Chase

JPMorgan has developed blockchain infrastructure for institutional settlement and provides crypto-related services to large clients. The bank has also been involved in digital asset custody development and continues to expand blockchain-based financial rails.

Citigroup

Citigroup has focused on institutional crypto services, including digital asset custody frameworks and tokenization initiatives designed to integrate blockchain technology into traditional capital markets.

While still described as “early journey” for yield/lending, Oldenburg’s comments make clear that Morgan Stanley views these services as logical extensions of its digital asset roadmap.

The bank’s expansion highlights a structural shift in how major financial institutions view Bitcoin. Rather than treating digital assets as peripheral investment options, banks are building long-term infrastructure to support ownership, financing, and income generation

For the Bitcoin ecosystem, the message is clear, the world’s largest financial institutions are no longer just observing, they’re building the on-ramps.

The Tekedia EDIA Play Framework

0

Every business is constructed on three foundational pillars: people, processes, and tools. These are the engines through which products and services are created to remove frictions in markets.

When talent is properly aligned, processes are intelligently designed, and tools are optimally deployed, firms create value at scale. At Tekedia Institute, our programs are designed to help co-learners master how to optimize these pillars for maximum market impact. In essence, we help organizations solve the calculus of the market where growth, friction reduction, and profit are the desired outputs.

To operationalize this thinking, we conceptualize market strategy through what we call the Tekedia EDIA Play framework:

  • Efficiency Play – Doing the same thing faster, cheaper, and more reliably. This is the domain of operational excellence, automation, and lean supply chains.
  • Differentiation Play – Making customers perceive you as meaningfully distinct through brand, design, superior experience, or specialized knowledge.
  • Innovation Play – Creating something truly new, shifting the basis of competition and redefining the rules so competitors must respond to your architecture.
  • Aggregation Play – Using scale, platforms, and networks to coordinate fragmented demand and supply, capturing value through orchestration rather than direct production.

Join me tomorrow at Tekedia Mini-MBA as we examine the Grand Playbook of Business and unpack how these four plays shape market leadership.

Tekedia Mini-MBA Edition 19 Has Started, Register and Join | $170 or ?120,000