ServiceNow AI Agents
ServiceNow AI Agents: Enterprise AI platform or workflow agent for knowledge management, sales, support, data, and organizational automation.
ABAB Structured Brief
ServiceNow AI Agents is indexed in ABAB Crypto Map under AI Models & Apps. This page keeps the official site, category, tags, and related ABAB coverage together as a searchable crypto project profile. Official domain: servicenow.com.
Related News & Analysis
Writer Dan Koe: Those Who Are Easily Addicted Possess One of the Rarest Traits
...t those who are easily addicted possess one of the rarest traits, as most people cannot maintain interest in anything for more than a week, while those prone to addiction can unknowingly invest 6 hours. Currently, this t...
Vurt Bets on Next Wave of Vertical Screen Entertainment Growth Coming from Wide-Screen TVs, Not Just Mobile
...mostly produced with live actors and crews, rather than the AI-generated content common among competitors, and allows free viewing through advertisements. Co-founder Hilmon Sorey stated that users initially wanted ...
Elon Musk Claims Address Has Been Exposed Multiple Times and Is Not a Residence
Elon Musk responded in a post, stating that a certain address has been publicly exposed multiple times and that he does not reside there; the address is a location where his security team receives mail. This response was...
Elon Musk Recruits World-Class Chip Engineers for SpaceX or Tesla
...hines. The post referenced discussions about Terafab, aiming to complete about 10 chip designs per year, while the industry typically completes one significant chip from design to mass production every two years.</...
Farewell to the Era of Simply Writing Code: YC Partner Discusses How Experienced Entrepreneurs and Independent Founders Use AI to Break the Deadlock
1. Core Data and Macro Changes in Venture Capital • Surge in Hard Tech Proportion: Among the projects selected by YC in the past 12-18 months, the proportion of hard tech (projects involving the physical world/atoms) has jumped from 8% to 20%. • Distribution of Sub-sectors: Robotics has risen from 1% to 6%-7%; domestic industrial manufacturing has climbed from 4% to 10%; defense technology has increased from 1.5% to about 5%; semiconductors and photonics are close to 4%; energy and computing power infrastructure has risen to nearly 3%. • Change in Founder Educational Background: In the current summer batch, 1 in every 6 founders holds a PhD, indicating a significant increase in the demand for a strong research and academic background in deep tech fields. • Business Launch and Revenue Growth Break Historical Records: • Previously, YC teams often had zero revenue upon selection, with a median monthly recurring revenue (MRR) of about $8,000 at graduation; now this median has surged to $20,000. • There have even been extreme cases where revenue broke one million dollars (Seven Figures) within just three months of the batch, whereas traditionally, a similar scale would typically require over 18 months. • Proportion of Solo Founders Soars: The proportion of solo founders in selected teams has skyrocketed from 5% to 18%-19%, setting a new historical high. 2. Underlying Drivers of the Hard Tech (Atoms) Explosion • Generative Code Eliminates Engineering Bottlenecks: Previously, hard tech startups faced stringent limitations such as supply chain issues and complex hardware-software integration, with top software engineers being a key scarce resource. AI-assisted coding and code generation have significantly lowered this barrier, allowing startups to advance full-stack hardware development without needing to hire large engineering teams. • Cutting-edge Models Feed Back into Scientific Breakthroughs: AI not only changes software but also significantly accelerates fundamental scientific research, enabling deep tech startups to achieve technology validation at earlier stages. • Three Major Macro Driving Factors: • Successful Examples of Space Commercialization: Breakthroughs by benchmark companies like SpaceX have driven investment in space infrastructure, leading to alternatives for sovereign satellite networks (such as Exosat) and orbital energy solutions (like Beyond Reach Labs developing satellite solar power for space data centers). • Geopolitical and National Defense Security Demands: The new generation of founders is directly entering the defense sector, breaking through the traditional cost-plus consulting model of the military-industrial complex. For example, Icarus is developing solar-powered high-altitude reconnaissance/communication aircraft, and Nine Mothers is developing computer vision-driven short-range defense systems against drones, both quickly securing lucrative military contracts. • Restructuring of Domestic Supply Chains and Dual-Use: Companies like Nox Metals are revitalizing old factories in Detroit to rebuild the domestic metal supply chain, with revenues growing at an ultra-high compound growth rate similar to software SaaS, primarily driven by the immediate procurement needs of emerging defense tech companies. 3. Evolution of Computing Power and Infrastructure: Chips, Optical Interconnects, and Power • Supply-Demand Reversal in Computing Hardware: Even the rental cost of older model GPUs (like A100) has seen a rare upward trend, driving the entire industry chain to expand, from data center site construction and power battery hybrid solutions to underlying processors. • New Architectures and Low-Precision Computing: The requirements for floating-point precision in large models continue to decrease (from FP32, FP16 down to FP8 and even lower), prompting teams exploring alternative solutions with ternary architectures or custom architectures (like Bot, Lamb Labs). • Optical Interconnects Break Network Bottlenecks: The rapid surge in GPU computing power has made traditional electronic switches within data centers a communication throughput bottleneck. Companies like Dipole Labs are beginning to develop all-optical switching systems to achieve end-to-end all-optical communication. 4. Reconstruction of Software and SaaS: From Tools to End-to-End Agents • Software is not disappearing but undergoing a qualitative transformation: The proportion of companies providing "end-to-end task completion" in the batch has increased from 10% to over 25%. • From Point Recording to "Direct Work": • Traditional SaaS is essentially point solutions or record systems that require human manual operation; • The new generation of products directly replaces complete workflows (such as full-process insurance brokerage, clinical appointment pre-registration, medical billing settlement, etc.), with customers more willing to pay a premium for "final results". • Example (Juicebox): Initially screening talent profiles through LLM, it has upgraded to having Agents automatically complete bulk outreach, schedule coordination, and preliminary interview arrangements, leading to a doubling or even several-fold increase in average revenue per user (ARPU). • Data Silos and AI Harness Competition: • Existing system record giants (like Salesforce, Slack) are evolving into AI harness frameworks. Companies with core workflow data and collaborative relationships can solidify their moat if they can make their software the main battlefield frequently called upon by Agents. 5. Invisible Money-Making Tracks: Laboratory Data and Reinforcement Learning (RL) Environments • Quietly Profitable Niche Markets: In the past two years, YC has invested in dozens of startups selling high-quality training data or reinforcement learning (RL) simulation environments to leading AI laboratories, with several teams established only a few years ago achieving annual revenues exceeding tens of millions or even hundreds of millions of dollars. • Capital Expenditure Flow: Leading large model laboratories are estimated to spend billions of dollars on RL environments and custom data construction, focusing on areas like financial deep simulation and long-horizon tasks. • Extending to the Physical World: Laboratories need to train embodied intelligent models, driving rapid revenue generation for companies collecting industrial operational network data, first-person perspective (Egocentric), and tactile/operational data (like Practis Robotics, Deep Reach, Human Archive, etc.). 6. Robotics Technology Paradigm: Fine-tuning and Specialized Data are Key • Complexity of Real-World Dimensions: Language models (LLMs) abstract the real world into language symbols, while robots need to handle continuous degrees of freedom in 3D physical space, requiring extremely high real-time responsiveness and physical interaction fault tolerance. • General Base + Vertical Scene Fine-tuning: Startups accessing general robotic base models (like Physical Intelligence's PI model) generally cannot be used out of the box and must combine proprietary video and action data from vertical scenes for deep fine-tuning (for example, Boost Robotics focuses on cable plugging and unplugging in data centers, while Ultra combines thousands of hours of packing data to delve into packaging flows). 7. Evolution of Founder Portraits: The Dividend Period for Experienced Practitioners and Solo Teams • Experienced Founders Welcome Spring: Entrepreneurs in their late 30s, 40s, or even over 50 have significantly expanded advantages. • Business Acumen Surpasses Pure Coding: After Agentic Coding has greatly reduced software construction costs, knowing "what to build" and having a deep understanding of vertical business pain points and industry insights have become the most scarce core barriers. • Agent Management Equals Team Management: Seasoned practitioners with rich engineering team or corporate management experience show stronger control when breaking down complex tasks and scheduling multiple concurrent coding Agents. • The Threshold for Solo Entrepreneurship Has Been Greatly Lowered: • In the past, running a company solo required extremely rare all-round talent (with top-notch technical, business, and sales abilities); now, with the help of AI, the threshold for a single person to prototype and validate commercialization has been significantly lowered. • Dynamic Partner Structure: The modern model tends to have a solo founder start and gain early market traction, and then introduce full-time partners and teams based on business needs at a more mature stage.
a16z Closed-Door Debate: From Existential Risk (X-Risk) Regulatory Traps, Internal Network Agent Protection to New Paradigms in Probabilistic Programming
1. Core Controversy: "Pacing" or Regulatory Capture? • The open letter from Anthropic founder Dario Amodei has sparked intense debate in the industry: • The frontier large model safety and sandbox isolation initiative proposed by Dario and others is reasonable at the pure technical engineering level; • However, the focus of the controversy lies in the narrative packaging of "pacing" and the so-called "existential risk (X-Risk)". • The logical flaw of the concept of "pacing": • Lack of a reference frame: There has never been a publicly agreed benchmark speed for technological evolution. Announcing "we have decided to slow down" when no one knows the original completion node is essentially similar to the media claiming "the unannounced Apple car has been delayed," which lacks measurable standards for a self-consistent narrative. • A middle ground that pleases neither side: Attempting to walk a compromise between the internal extreme pause faction (Pause/Doomer) and external regulatory bodies has resulted in radicals accusing it of "just slowing down instead of stopping," while regulators deem it as "acknowledging the existence of harm yet still racing ahead." • Reflecting on the moral coercion of "existential risk (X-Risk)": • Historically, nuclear weapons development teams were well aware of their absolute lethality on a physical level, thus establishing the highest level of national control mechanisms; • If the heads of large model laboratories genuinely believe there is a 10% probability of human extinction, the only ethical response would be a complete halt or total nationalization; • The reality is contradictory: If one side portrays extinction risk while continuing to accelerate financing and pushing into the commercial market, it can easily be exploited by politicians, leading to excessive regulatory capture that could strangle startups and the open-source ecosystem. 2. Historical Reflection: From Nuclear Bombs, Early Internet Viruses to Regulatory Lag • Prerequisites of the "Fact Pattern" in policy-making: • The automotive industry was already widespread in the early 20th century, and it wasn't until Ralph Nader published "Unsafe at Any Speed" in the 1960s that stringent safety regulations were formed; • The aviation industry experienced a 40-year period of technical trial and error from the Wright brothers' first flight to the establishment of a complete FAA airworthiness certification system; • Traditional regulation must be based on specific damages that have occurred and a clear causal chain; attempting to implement "predictive legislation" before technology has matured is almost certain to stifle innovation. • If we were to apply today's fear paradigm to the Internet of the 1990s, it would never have emerged: • In the era of Windows 95 and early PCs, connected computers faced widespread risks from worm viruses and network interruptions (for example, the Morris worm paralyzed 10% of backbone networks, leading to frequent disconnections and business losses); • At that time, Congress passed the Computer Fraud and Abuse Act (CFAA) targeting specific system intrusion cases; if the Internet's underlying infrastructure had been completely locked down in 1994 out of fear of hackers, the modern digital economy would not exist. • The potential spread risk of the European GDPR model: • Europe, lacking local underlying tech giants, tends to legislate aggressively in compliance and antitrust areas; • Beware of Europe refining regulation into a "digital airbag reminder" for AI Agents—where every time an Agent calls an external API or performs file read/write, a GDPR-like disclaimer confirmation box pops up, ultimately leading to widespread user cognitive numbness. 3. The Real New Security Front: Engineering Threats from the Emergence of Internal Network Agents • The complete failure of traditional internal network security assumptions: • Past IT and enterprise information security were based on an implicit assumption: 95%-99% of internal employees comply with regulations most of the time, with malicious insiders being a very low probability event. • Internal GitHub, Slack, approval flows, and financial reimbursement systems often lack strict concurrency throttling, relying solely on single sign-on (SSO) and coarse-grained identity authentication. • Agent swarms as "internal network distributed denial-of-service attacks (DDoS)": • When internal employees start to batch schedule thousands of autonomous Agents to write code, call APIs, and run automated tests, the software behavior is structurally identical to high-frequency DDoS attacks. • Agents possess rapid retry and tireless characteristics, easily exhausting internal microservice resources in a dead loop, and may even trigger dangerous data overwrites due to misunderstanding instructions. • Next-generation operating systems and permission granularity innovation: • The existing operating system permission system is too crude (either fully open or frequently prompting for confirmation); • There is an urgent need to reconstruct the underlying security stack to support fine-grained dynamic permission isolation (e.g., instantaneous read/write control for specific folders, adaptive API rate monitoring, and audit tracking), shifting the security focus from the metaphysical level of species survival back to concrete engineering defenses. 4. Paradigm Shift: The Rise of the Jev Model and the Return of Probabilistic Programming • The essential conflict between natural language interaction and traditional software systems: • The industry has previously focused on a "text input, text output" chat model, but forcing structured software to parse unstructured long text is extremely costly and error-prone; • Relying on complex prompt constraints and schema validation cannot guarantee 100% deterministic output and wastes enormous computational costs and response delays on meaningless question-and-answer redundancy. • The core breakthrough of the Jev model: from "generating text" to "semantic decision-making": • The input remains semantic context, but the output completely abandons generative text; • Among a given set of discrete options or routing conditions, it directly returns the optimal decision and precise probability distribution with extremely high throughput and low latency; • This completely overturns inefficient chat box interactions, allowing traditional software code to directly use large model outputs as the basis for logical branching decisions. • The revival of half a century of computer science legacy: • The core proposition of programming languages in the early 1960s-1970s was simulation and probabilistic modeling (e.g., handling ballistic trajectories, wind fluctuations, and other non-deterministic physical systems); • Modern software code has long been limited to Boolean algebra (absolute certainty of if/else); • With the implementation of Jev-like architectures, the core programming paradigm is shifting entirely to condition branches based on probabilistic confidence (if x% then ...), seamlessly integrating probabilistic language models with classical deterministic software engineering. 5. The Shift of Innovation Focus: From the Base Model Layer to the Model Periphery • The "critical mass dilemma" at the giant platform layer: • Leading large model laboratories are currently mired in maintaining large infrastructure operations, resolving alignment disputes, and managing massive user compatibility, severely dispersing their focus, making it difficult to optimize all vertical scenarios from the platform layer. • The autonomous evolution of application layers and peripheral tools: • The history of the software industry shows that core innovations often emerge outside the platform (analogous to the "Sherlocking" process where third-party independent tools are integrated into systems); • Large models are no longer merely revered as ultimate deities but are evolving into standard underlying components like databases and compilers; the core technological moat of future software is shifting entirely to how to reconstruct contextual systems, state persistence, and high-reliability business flows around the model's periphery.
Valuation of $9.5 billion: Interview with Gusto co-founder on returning to the front lines to write code and create a disruptive "AI co-founder" for logistics.
1. Prelude to Entrepreneurship and Opportunities: From Solo Development to Billion-Dollar Pain Points • The "Side Hustle Ceiling" of Solo Developers: • Before founding Gusto (formerly ZenPayroll), Eddie had shut down his first startup and faced a brief financial struggle. • He then self-learned and keenly seized the early ecological benefits of the Android system, developing several pure tool applications (such as Car Locator, which solved the problem of finding cars, and Screenshot It, which was launched when the system did not natively support screenshots, charging as much as $7 for a single app). • At his peak, his personal annual income approached $700,000, achieving financial and time freedom, but due to the keen awareness that the underlying operating system would eventually consume these lightweight single-point tools, he decided to seek a more serious venture with long-term compounding and social value. • Serendipitous Opportunity and Meeting of Founders: • With no marathon training, Eddie impulsively participated in the San Francisco Half Marathon, where he coincidentally met his classmate Josh Reeves from Stanford University's Electrical Engineering Department at the starting line. • After the race, the two had coffee at The Creamery (a well-known early project and venture capital gathering place in Silicon Valley) at the intersection of Fourth Street and King Street, discovering that they both desired to create substantial products that addressed profound real-world problems. • They later introduced Tomer London, a high-achieving student from the Technion, through the Stanford alumni network, forming a co-founder team of three with engineering backgrounds. • The Origin of Underlying Business Insights: • Eddie's parents ran a private clinic in Southern California during his childhood, with his father handling patient reception at the front desk while his mother took on all the tedious logistical support (scheduling, payroll, tax reporting, reimbursements, employee social security, medical compliance). • His mother's daily complex labor on non-core business tasks became a seed of pain buried deep in Eddie's memory, ultimately giving rise to Gusto, a payroll management platform that automates core processes for small and medium-sized enterprises (serving over 300,000 small businesses, valued at over $9.5 billion). 2. The Psychological Journey and Self-Management of Early Technical Co-Founders • The Psychological Imbalance of "Coding vs. Eating Steak": • Many technical founders (Founding CTOs) often experience psychological imbalance in the early stages—spending 16 to 18 hours a day coding at the computer, while their business/CEO partners, due to the early business not yet being underway, mainly spend their time having coffee with investors, watching games, and eating steak to raise funds. • Breaking the Cognitive Barrier: This is not unfair but is determined by the different roles within the startup team; one must completely let go of the defensive "ego" and look at the long term, as the pressures and burdens borne by different partners will dynamically balance at their respective stages. • Clarifying Early Roles and Ownership: • After being selected for Y Combinator (2012 batch, Eddie had previously participated in early YC in 2008), the three founders sat around three chairs and had an extremely candid and even uncomfortable conversation, establishing on the spot that Josh would serve as CEO, Tomer would be responsible for product (CPO), and Eddie would handle technology (CTO), avoiding future control disputes that lead to 70% of startup internal conflicts. • From IC (Independent Contributor) to Team Manager: • When the team expanded to 10 engineers, Eddie often delayed one-on-one meetings with employees due to being engrossed in fixing specific bugs, leading to management process bottlenecks. • After being pointed out candidly by the CEO, he consulted senior advisors like former Dropbox CTO Aditya Agarwal and former VMware CTO Steve Herrod, systematically transforming into a full-time technical manager, completely saying goodbye to coding after 10 years. • Physiological Coping Techniques for Emotional Triggers: • When feeling attacked or challenged, the human body spontaneously enters a "fight or flight" state (breath holding, accelerated heartbeat), and the brain's higher rational decision-making areas temporarily shut down. • Practical Approach: Recognize one's bodily reactions (such as noticing oneself holding their breath), force oneself to "hit the pause button," excuse oneself to wash hands in the restroom for a few minutes, create physical distance to reset emotions, and shift from passive defense to an active choice of a growth perspective. 3. Returning to the Front Line: The Transformation of a Technology Manager Driven by Large Models • The Catalyst for Resuming Coding and the Heathrow Airport Encounter: • In early 2026, Eddie was exposed to open-source and cutting-edge AI experiments (such as the Telegram-driven OpenClaw and other personal agent environments), and was shocked by the concept of "periodic active polling and automatic triggering (Heartbeat mechanism)." • In February 2026, during a return trip from Madrid, he was stranded at London Heathrow Airport for 5.5 hours due to a delayed previous flight. While sitting in the VIP lounge, Eddie used cutting-edge tools like Claude Code to independently write a groundbreaking prototype system in just 5 hours. • AI Bridging the Gap in Technology and Learning Shame: • Over the past 10 years, due to focusing on HR and strategic management, Eddie became disconnected from modern enterprise-level toolchains (for example, he had never personally configured DataDog or built complex data dashboards), and returning to architectural development carried a significant psychological burden. • AI programming assistants became a "top-tier technical mentor without any disdain or judgment bias," allowing managers to bridge a 10-year technology stack gap within days, independently producing production-level prototypes without involving any engineering subordinates. 4. The New Product "Co-founder": Reshaping the Full Workflow for Small and Micro Enterprises • From Pure Applications to "Proactively Executing Autonomous Workflows": • The initial concept was to allow small business owners to generate their own small CRUD web applications through dialogue on the Gusto interface, similar to using Lovable; • However, after in-depth communication with users, it was found that the fundamental pain point for small and micro enterprises was not "lacking an interface," but rather "lacking people to thoroughly complete the miscellaneous tasks." • Product Essence: Users communicate their business logic to "Co-founder" through daily communication channels (SMS, WhatsApp, Slack, email) in natural language, and the AI agent autonomously completes the fulfillment end-to-end according to the set cycle (Heartbeat). • Realistic Typical Scenario: Payroll Process Replacement for Offline Massage and Wellness Centers: • In the traditional model, store owners use scheduling systems (like MindBody) to manage schedules, including fragmented data such as technician service duration, different commission ratios for essential oil products, different project bonus distributions, and tips for each technician; • Owners typically need to first export raw data from the scheduling system to Excel, write various functions to break down performance and bonuses, and after calculations, manually input the results into Gusto for distribution. • Co-founder's takeover solution: The AI agent directly calls underlying tables or system data sources, automatically parses extremely vertical and complex personalized commission rules, and directly completes payroll verification and final approval in the background, reducing what used to be hours of desk work to zero. 5. Organizational Efficiency Transformation: "AI Builders" and Rapid Delivery in Ten Weeks • Saying Goodbye to Jira, Scheduling Planning, and Bureaucracy: • The agile core team for building Co-founder consists of only 5 to 8 people, collectively referred to as "AI Builders," completely abandoning traditional long-term technology roadmaps, cumbersome Jira task breakdowns, and formalistic meetings. • Cross-Functional Integration: Product designers in the team, aided by AI, can now directly write and submit high-quality production-level front-end and full-stack code to the main codebase, breaking the inefficiency of "producing Figma prototype designs and then handing them over to engineers for secondary translation." • Extreme Implementation Cycle: • In early March, initial whiteboard architecture was conducted in Denver; • Five weeks later, it was delivered in person to 10 seed customers in New York for practical validation and feedback collection; • Another five weeks later, it opened public testing and Waitlist for global customers. The entire process took only 10 weeks, achieving a speed of implementation akin to a hacker marathon within a large company. • Core Insight for Modern Technical Leaders: • Whether at the founding level or among mid-to-senior technical managers, at this current turning point of technological qualitative change, one must return to the ground and get hands-on (Hands-on). • Relying solely on subordinate reports or macro dashboards cannot perceive the architectural bottlenecks in today's real software development processes, while personally writing code can simultaneously reshape the three-dimensional aspects of real customer needs, developer engineering experience, and the underlying product moat.