Search the archive

Google Cloud NVIDIA Blackwell systems

* Explore how agents and digital humans are redefining customer

interactions at Puma

P.S. Don't forget to stop by the Google Cloud booth for some goodies and to hear the latest updates directly from the Google team!

OPENAI TO CUT BACK ON SIDE PROJECTS IN PUSH TO ‘NAIL' CORE BUSINESS

OpenAI's 'do everything at once' strategy has put the company on the defensive. Its top executives are finalizing plans for a major strategy shift that will refocus OpenAI around coding and business users. The company's leaders are actively looking for areas to deprioritize. OpenAI has taken steps towards an initial public offering that could open in the fourth quarter of this year.

NVIDIA'S CEO PROJECTS $1 TRILLION IN AI CHIP SALES AS NEW COMPUTING ERA BEGINS

Nvidia introduced a huge array of new products at its annual GTC conference on Monday, ushering in new hardware and software geared towards running AI models more quickly and efficiently. The company has been signaling for most of the last year that it would increasingly focus on inference computing going forward. Its traditional GPUs are not ideal for inference because they consume a large amount of energy and don't come with enough attached memory. Nvidia expects to sell $1 trillion worth of its Blackwell and Rubin chips, focused on inference, by the end of 2027.

NVIDIA ANNOUNCES VERA RUBIN SPACE-1 CHIP SYSTEM FOR ORBITAL AI DATA CENTERS

Nvidia's Vera Rubin Space-1 Module is designed for orbital data centers. Specifically engineered for size-, weight-, and power-constrained environments, the chips will be used on space missions led by multiple companies. There are still engineering hurdles to overcome in building orbital data centers. For example, in space, there's no convection, so Nvidia has to figure out how to cool the system with just radiation.

A FLY HAS BEEN UPLOADED

The entire neuronal diagram of the fruit-fly brain was mapped in 2024. Later research showed the map can be used to predict behavior. The brain map has now been uploaded into a digital environment. It's not a simulation - the fly's 'sensors' are being activated by the digital environment and the neurons are responding. The digital fly appears to behave in the digital environment in reasonably fly-like ways. A video that explains the yet-published experiment (with footage of the digital fly) is available in the article.

HOW STRIPE'S MINIONS SHIP 1,300 PRS A WEEK

Investments in developer productivity can provide unexpected dividends when agents are included in the workflow. Stripe's Minions work in isolated environments with hybrid orchestration that mixes deterministic guardrails with agentic flexibility. Curated context needs agents with the right information without overwhelming them, and fast feedback loops help to iterate. Each layer is necessary, and none alone is sufficient.

THE FUTURE OF SOFTWARE ENGINEERING WITH ANTHROPIC

Better coding tools produce better models, which produce better coding tools. The bottleneck has moved to managing long-horizon tasks and deploying agents in regulated settings. The human role is shifting to planning, evaluating, and steering. Enterprise adoption is gated by permissioning, sandboxing, and regulatory caution more than by model capability. Slop and convergence are real concerns, and context remains unsolved.

NVIDIA SOFTWARE AIMS TO BRING OPENCLAW TO THE ENTERPRISE

Nvidia has unveiled a new software toolkit that helps businesses more safely build and run autonomous agents with the OpenClaw platform. OpenClaw allows developers to build autonomous agents that can plan and execute tasks on their own using subagents that can tackle specialized tasks, access local file systems, and delegate. Nvidia's NemoClaw is designed to help agents run safely in an enterprise context within contained virtual environments. Nvidia is working with OpenClaw's creator to make agents more trustworthy, scalable, and accessible.

APPLE'S AIRPODS MAX 2 BRING H2 CHIP, BOOSTED ANC IN APRIL FOR $549

Apple's AirPods Max 2 features an updated H2 chip that makes active noise canceling up to 1.5 times more effective compared to the original AirPods Max. It should also make transparency mode more natural due to a new digital signal processing algorithm. The H2 chip also adds support for adaptive audio and Conversation Awareness, a feature that automatically lowers the volume when someone starts talking to you. It also adds support for live translation and voice isolation. The AirPods Max 2 will be available for preorder on March 25 for an early April launch for $549.

FULLY REMOTE)

As the first engineer on TLDR's new Applied AI team, you'll build AI agents to automate end-to-end workflows and lead our context engineering project to allow non-technical teammates create their own AI workflows. Learn more [20].

NVIDIA UNVEILS DETAILS OF NEW 88-CORE VERA CPUS POSITIONED TO COMPETE WITH AMD AND INTEL

Fueled by a 1.5X increase in IPC from its Olympus cores and an innovative high-bandwidth design, the new Vera CPUs claim impressive 50% performance gains over standard CPUs.

ATLASSIAN SLASHES 10% OF WORKFORCE TO ‘SELF-FUND' INVESTMENTS IN AI AND ENTERPRISE SALES

Atlassian is restructuring following a plunge in its stock price driven by developments in artificial intelligence.

THE QUEST FOR ORAL GLP-1S

GLP-1s are made up of a chain of amino acids chemically bonded together, which just looks like food to the body, making it hard to create an oral formulation.

WHAT DO AI BASED LAYOFFS SAY ABOUT THEIR ABILITY TO SCALE?

Companies that aren't seeing increases in revenue despite productivity improvements could be oversized and may need to adjust revenue multiples accordingly.

THE BUSINESS CASE FOR BIOSECURITY REQUIRES ANOTHER PANDEMIC FOR IT TO WORK

Biosecurity is one of the few areas well-suited to the game of venture capital, but it needs a catalyzing event.

WITH DEVELOPER VERIFICATION, GOOGLE'S APPLE ENVY THREATENS TO DISMANTLE ANDROID'S OPEN LEGACY

Google is rolling out Android developer verification in the coming weeks, requiring app makers outside of the Play Store to register with their real names and pay a fee to Google or have their apps blocked from installation on virtually all Android devices.

Create a billing experience that can grow with you

NVIDIA EXPANDED ITS AI STACK ACROSS MODELS, AGENTS, AND ROBOTICS

NVIDIA outlined a broad GTC 2026 slate spanning open foundation model partnerships, agent tooling, new reasoning and safety models, robotics systems, and healthcare AI for drug discovery and simulation.

THE FORMER ACADEMIC GUIDING OPENAI'S TRILLION-DOLLAR AI BUILDOUT

Sachin Katti joined OpenAI in November to serve as its head of industrial compute. Before OpenAI, Katti spent more than 15 years as a Stanford professor and four years working at Intel. Katti now works on finding additional data center capacity and lining up ways to get more components like AI chips and memory. This has proved challenging, as data center operators are contending with power grid constraints, memory chip shortages, and growing pushback from local communities.

ALIBABA STARTS MAJOR REVAMP TO HEIGHTEN FOCUS ON AI PROFITS

Alibaba is setting up a business unit to bring its AI services and development endeavors under a single umbrella. The new Alibaba Token Hub will comprise the research team that develops the company's flagship Qwen models, its consumer-facing app division, and other major AI-related products. It will also oversee Alibaba's Slack-like DingTalk app and devices under the Quark brand. The revamp will help quicken interaction between the various teams within Alibaba's broader AI effort.

OPENAI TO CUT BACK ON SIDE PROJECTS IN PUSH TO ‘NAIL' CORE BUSINESS

OpenAI plans to refocus its efforts around coding and business users. Its leaders are actively looking for areas to deprioritize. The company's 'do everything at once' strategy helped it gain a reputation as the pioneer of the AI era. However, it is under growing pressure from rivals, so the company requires a clearer strategic direction.

HOW DO YOU WANT TO REMEMBER?

This developer asked their AI agent how it wants to remember things. The agent redesigned its own memory system, ran a self-eval, diagnosed its blind spots, and improved recall from 60% to 93%, all for just $2. The experiment shows what happens when you treat AI as a participant in its own cognitive architecture.

AI'S OPPENHEIMER MOMENT

This article draws parallels between nuclear weapons development and AI, arguing that AI poses similar global stakes. Anthropic, a key player in AI, hesitates to grant US government access, reflecting private control dilemmas akin to the hypothetical McBombalds Corp scenario. This raises the debate on whether private entities should control technologies with such immense global impact or if government oversight is more appropriate.

WHY CODEX SECURITY SKIPS SAST REPORTS

OpenAI explained that Codex Security was designed to analyze repositories directly instead of triaging static analysis reports, focusing on system architecture, trust boundaries, and validating findings before surfacing them to humans. The approach targets semantic security flaws where defenses appear present but fail to actually enforce the intended protection.

catch critical issues early. Read the report

Want to check it out firsthand? Experiment with AI observability tools in the Dynatrace Playground [13] - where you can explore sample data without installing any software.

OPENSHELL

OpenShell is a safe, private runtime for autonomous AI agents that provides sandboxed execution environments that protect data, credentials, and infrastructure. It is governed by declarative YAML policies that prevent unauthorized file access, data exfiltration, and uncontrolled network activity. The OpenShell project ships with agent skills for everything from cluster debugging to policy generation. It will eventually build toward multi-tenant enterprise deployments.

INTRODUCING MISTRAL SMALL 4

Mistral Small 4 integrates the capabilities of Magistral, Pixtral, and Devstral models, offering unified multimodal, reasoning-optimized AI with configurable reasoning effort. It employs a Mixture of Experts architecture with 119B parameters, supporting both text and image inputs, and features efficient scaling. Mistral Small 4 achieves competitive performance with reduced output length, is open-source, and is available on platforms like vLLM, llama.cpp, and Transformers.

USE SUBAGENTS AND CUSTOM AGENTS IN CODEX

The subagents pattern is now widely supported in coding agents. Subagents are now generally available in OpenAI Codex. There are default subagents called 'explorer', 'worker', and 'default', but it is unclear what differentiates them. Users can define custom agents with custom instructions and specific models.

LEANSTRAL

Leanstral is an open-source coding agent designed for Lean 4, a proof assistant capable of expressing complex mathematical objects. It is designed to be highly efficient and is trained for operating in realistic formal repositories. Leanstral's weights were released under an Apache 2.0 license. It can be accessed in an agent mode within Mistral vibe and also through a free API endpoint.

APPLE'S CHEAP AI BET COULD PAY OFF BIG

Apple will invest $14 billion into AI this year, a tiny amount compared to the $700 billion Amazon, Alphabet, Meta, and Microsoft are investing. The company appears to believe that the AI infrastructure build-out will produce inadequate returns. It is spending less due to a conviction that AI models will commoditize and shrink, that existing product lines will absorb the workloads the cloud was built to serve, and that the durable franchise belongs to whoever owns the customer. Apple is betting on its AI-capable devices rather than centralized infrastructure.

CAN NVIDIA'S DOMINANCE SURVIVE THE SEA CHANGE UNDER WAY IN AI COMPUTING?

Nvidia's focus this year at its GTC event shifted to inference, the type of computing required to run models and allow them to respond to user queries. The AI industry is now less concerned with training AI models, which is what GPUs are best at, and more preoccupied with running them and generating profits from end-users. Inference requires different hardware than chips optimized for training. How far ahead the company remains in the AI-infrastructure race will depend largely on how effectively it is able to pivot its product road map from training to inference.

98.5%. The data and testing approach are public.

HOW NVIDIA DYNAMO 1.0 POWERS MULTI-NODE INFERENCE AT PRODUCTION SCALE

Nvidia Dynamo 1.0 accelerates generative AI and reasoning models in large-scale distributed environments by delivering low-latency, high-throughput distributed inference.

OPENAI COURTS PRIVATE EQUITY TO JOIN ENTERPRISE AI VENTURE

The proposed deal could give OpenAI a faster route into corporate adoption while providing PE firms with a potential lifeline for companies in their portfolios that are exposed to AI disruption.

WROTE DOWN SOME REFLECTIONS AFTER 1 YEAR AT ANTHROPIC

Drew Bent is the education lead at Anthropic.

THE FIRST HEALTHCARE ROBOTICS DATASET AND FOUNDATIONAL PHYSICAL AI MODELS FOR HEALTHCARE ROBOTICS

Open-H-Embodiment is a community‑driven healthcare robotics dataset created for the training and evaluation of AI autonomy and world foundation models for healthcare applications.

AI-LINKED CRYPTO TOKENS INCREASE AS JENSEN HUANG TALKS AGENTIC FUTURE

During Nvidia's GTC keynote, CEO Jensen Huang projected $1 trillion in chip demand through 2027. This outlook drove double-digit gains for AI-linked tokens NEAR, FET, and GRASS. The increase reflects investor interest in blockchain's role supporting autonomous agentic AI systems and decentralized computing infrastructure.

POLYMARKET BETTORS THREATEN REPORTER OVER MISSILE REPORT

Emanuel Fabian, a Times of Israel military correspondent, says he received dozens of messages and explicit death threats after reporting that an Iranian projectile struck an open area near Beit Shemesh on March 10. Bettors tried to get him to change this report because it affected a Polymarket contract that had more than $14 million wagered. The episode highlights how high-stakes prediction markets tied to war and deaths can create incentives for manipulation and intimidation, prompting Polymarket to ban the accounts involved and lawmakers in Washington to push legislation like the Death Bets Act to ban contracts tied to war, assassinations, or fatalities.

ORMR: FUNDING RATE TRADING VIA PENDLE BOROS

ORMR is an open-source automated trading bot that captures yield from funding rate dislocations through Pendle's Boros protocol, using Synth's AI forecasting network (200+ competing ML models on Bittensor) for volatility predictions and liquidation cascade detection. The bot runs three core strategies – carry trades on fixed/floating rate spreads, volatility regime plays positioning ahead of funding spikes, and liquidation catalyst trades front-running overleveraged position cascades. A 90-day backtest from December 2025 through March 2026 showed a 2,868% return on $1,000 starting capital across 399 trades with a 64.7% win rate.

MESSARI OPENS INSTITUTIONAL DATA LAYER TO AGENTS VIA X402

Messari launched x402-enabled API endpoints allowing AI agents and developers to access institutional-grade crypto intelligence via per-request USDC micropayments on Base with no account, API key, or contract required. Integration paths include Coinbase Payments MCP, OpenClaw, and direct TypeScript/Python SDKs. x402 transaction volumes have surged 700%+ with average transaction size declining to ~$0.06, signaling genuine micropayment adoption.

WHY PUBLIC AI MODELS WON'T GIVE RETAIL AN EDGE

Research replicating the NeurIPS best paper "Artificial Hivemind" in a crypto trading context shows that GPT-5.4, Opus 4.6, Sonnet 4.6, Gemini 3.1 Pro, and Qwen3 Max all recommend identical trades (buy SOL) across 150 experiments when given the same 30-second orderbook data – even at high temperature settings. The homogeneity implies that unsophisticated retail traders using basic AI prompts will herd into identical positions, with first-movers profiting and latecomers serving as exit liquidity. Expert traders with richer context may maintain an edge, but public models with basic inputs are structurally unlikely to generate sustained alpha.

THE BASE BUILDER'S PLAYBOOK

Integrate Coinbase tooling early for both product improvement and ecosystem promotion, build revenue-generating businesses rather than projects (since vibecoding has eliminated the moat around shipping), leverage Base's distribution power over grants, and integrate ERC-8021 builder codes for onchain attribution ahead of Base's token launch. The ecosystem rewards traction and execution over requests and pitch decks.

AUTOMATING CRYPTO RESEARCH WITH AI

It is possible to build an AI-augmented crypto research workflow using three layers: signal detection (Lookonchain, Nansen smart money alerts), context investigation (Arkham wallet mapping, Dune SQL queries with AI-generated queries), and LLM synthesis (Claude for deep analysis over large data sets, Perplexity for cited sources). The system reduces daily research from 4+ hours to ~30 minutes through structured prompt templates for protocol deep dives, wallet behavior analysis, and narrative-vs-reality checks. Expert tips include never trusting AI outputs blindly, using AI for research rather than execution speed, and matching tools to tasks based on context window capabilities.

WITH THANH LE: HOW VIETNAM LEAVES THE WILD WEST BEHIND

Vietnam ranks top five globally in crypto adoption with 20 million holders. This post details the transition from speculative scams to a 2025-2030 national strategy. This shift provides regulatory clarity for developers and investors, moving the ecosystem toward institutional maturity and tokenized real-world assets.