AI's Ripple Effect: Is the Digital Wave Cresting?

AI Exposure: A Statistical Mirage

The narrative of AI's pervasive impact on the workforce is beginning to fray. According to analysis from Gusto, a payroll and HR provider, the percentage of work exposed to AI has barely budged. From January 2023 to November 2025, this figure crept from 16.4% to a mere 16.3%, a statistically insignificant change that undermines claims of widespread AI adoption and its transformative effects.

This data from Gusto, which has deep visibility into small business employment, paints a stark picture. Despite the hype and vendor FUD, the actual integration of AI into daily workflows appears to be minimal, suggesting that the 'AI revolution' is largely confined to conference room buzzwords and investor pitches, rather than tangible shifts in how people work.

The limited impact is further evidenced by Gusto's own feature rollouts. While the company announced major updates in 2025, including S Corp support and 401(k) integration, AI-specific advancements weren't highlighted as major drivers of change in their recap. This suggests that AI, for many businesses, remains an add-on rather than a core component of their operational evolution.

Gusto’s recent updates, such as enhanced employee filtering via UUIDs and new include parameters for fetching employee data, released in June 2025, focus on essential HR functionalities, not revolutionary AI integrations. This practical approach to product development contrasts sharply with the grand promises made by AI evangelists.

The implication is clear: for a significant portion of the economy, particularly small and medium-sized businesses, AI hasn't yet delivered a compelling reason to fundamentally alter their operations. The promised efficiency gains and productivity boosts remain largely theoretical. OpenAI says its new model GPT-2 is too dangerous to release (2019) is a stark reminder of how fears of AI's power can overshadow the practical challenges of its implementation and adoption.

The Pragmatic Approach: Snowflake and Linear

While Gusto's data highlights a general lack of impact, examining specific company rollouts reveals a similar pattern of incrementalism. Snowflake, a major data cloud provider, has been steadily integrating AI capabilities, but these often feel like sophisticated enhancements rather than foundational shifts. The introduction of AI_COMPLETE function in late 2025 and Cortex Agents generally available in November 2025 are significant, but they extend existing data analytics and workflow automation rather than creating entirely new paradigms.

Similarly, Snowflake's February 2026 updates, including the CORTEX_AGENT_USAGE_HISTORY view, indicate a focus on monitoring and managing AI functionalities that are already in place. The GA of Cortex AI Agents in collaboration tools by Q1 2026 integrates AI assistants for data insights, a valuable but not revolutionary step. It’s about making data more accessible, not fundamentally changing how data is generated or utilized at its core.

Linear, a project management tool, has also embraced AI, launching 'Linear Agent' and touting 'Triage Intelligence & Automation' in 2026. This move positions Linear as an 'AI-powered product developer,' but the core functionality described—auto-creating issues from Slack—is an automation feature, albeit a smart one. The review from utilo emphasizes its evolution from a fast issue tracker to an AI-enhanced platform, but the fundamental nature of project management appears unchanged.

Linear's 'Now' updates also reflect a focus on core product improvements and customer success stories, such as the engineers at Base, rather than a radical departure driven by AI. While AI agents can streamline workflows, they do not fundamentally alter the nature of software development or project oversight, which have seen more profound shifts from version control to CI/CD pipelines. These advancements, while useful, appear to be optimizing the existing digital wave, not launching a new one.

The pattern across these companies—Snowflake, Linear, and even Gusto—is one of augmentation rather than revolution. AI features are being added to existing platforms to improve efficiency and streamline tasks, but they are not creating the foundational paradigm shifts that characterized previous digital waves like the internet, mobile, or cloud computing. As we explored in AI's Collision Course: Navigating Backlash Amidst Rapid Advancement, the practical integration poses significant challenges that temper revolutionary claims.

AI Benchmarks and Fleeting Dominance

The narrative of AI's inevitable dominance is heavily fueled by benchmark races and increasingly sophisticated, yet often narrow, demonstrations of capability. While techniques like fine-tuning, as detailed in our guide on Qwen3.5 Fine-Tuning Secrets, push performance boundaries, these often serve to validate the underlying models rather than signal a new wave of application. The constant pursuit of higher scores on leaderboards can create an illusion of exponential progress.

The rapid advancement in AI benchmarks, including those we've analyzed in How We Broke Top AI Agent Benchmarks, highlights the industry's focus on quantifiable metrics. Yet, these benchmarks often test AI in isolated, often artificial, environments. The real-world application for the average user or business may not see a commensurate uptick in utility or transformative capability.

The race for more powerful models, often discussed in the context of AI's future trajectory, can distract from the implementation challenges and the actual value delivered. For instance, while compact models are making strides in vulnerability discovery, as seen in Compact AI Models Now Rivaling Giants and Tiny AI Models Now Uncover Big Flaws Too, their application is still niche compared to the broad impact promised by overarching AI narratives.

The Market's Insatiable Demand for 'Next Big Thing'

The tech industry, perpetually seeking the 'next big thing' to drive investment and growth, has latched onto AI with an almost religious fervor. This creates a powerful echo chamber where incremental advancements are amplified into revolutionary shifts. The pressure to innovate and maintain market relevance often leads companies to frame even minor AI feature additions as groundbreaking.

This cycle of hype is not new. Recall the early days of AI research, where ambitious pronouncements often outpaced practical delivery. The concerns raised by organizations like OpenAI about the potential dangers of models like GPT-2 in 2019, while prescient about AI's power, ironically contrast with the current reality where demonstrating broad, tangible utility is the greater challenge. The perceived danger may have been overblown, or at least, premature.

The investment landscape further exacerbates this. Venture capital flows readily into AI startups, driven by the expectation of massive returns. This influx of capital can sustain companies and fuel development, but it can also lead to a proliferation of solutions that don't necessarily address fundamental market needs, creating a bubble of perceived innovation. This is a pattern we've seen before, as discussed in AI's FAAANGs: The Next Big Thing Is Already Here.

Defining the 'Digital Wave': What Comes Next?

The current digital age has been shaped by distinct waves: the personal computer, the internet, mobile computing, and cloud infrastructure. Each wave introduced new computing paradigms, fundamentally altering how we interact with technology and information. The PC brought computing to the masses, the internet connected them, mobile untethered them, and cloud centralized vast resources. These were not just feature updates; they were new foundational layers.

The pivotal nature of these past waves—the PC, the internet, mobile, cloud—lies in their ability to unlock entirely new categories of applications and businesses. They provided new substrates upon which innovation could flourish, leading to exponential growth and societal change. The question for AI is whether it represents such a fundamental shift or a mere optimization of existing digital substrates. As explored in AI's Core Revealed: Your Step-by-Step LLM Internals Guide, the underlying technology is complex, but its application layer is showing limitations.

AI as Optimization, Not Revolution

The features being rolled out by major tech players—Snowflake's Cortex Agents, Linear's AI Agent, and even Gusto's incremental updates—largely fit into the category of 'optimization.' They make existing processes more efficient, data more accessible, and workflows smoother. However, they don't fundamentally alter the user's relationship with computing or create entirely new digital realms in the way the internet or smartphones did.

Platforms like Enso are making autonomous agent deployment more accessible, but the use cases for these agents are still largely confined to automating tasks within existing frameworks. This is a critical distinction: powerful tools are being built, but they are being applied to refine the edges of the current digital wave, not to propel us into a new one. The focus is on making current digital activities better, faster, or cheaper, rather than enabling entirely new digital experiences or economies.

The trend towards local AI, as seen with frameworks like Ollama on Apple Silicon Apple Silicon Flees to Ollama: Is Your AI Already Obsolete?, and projects offering local RAG capabilities Local RAG Is Here: Your AI, Your Rules, No Cloud Needed, signals a desire for control and privacy, but also a potential acknowledgment that the centralized, cloud-based AI model might not be the singular path forward for a new wave. It suggests AI might be integrated incrementally into existing devices and workflows, akin to how specialized processors became common.

Computational and Data Bottlenecks

Despite advances in hardware and algorithmic efficiency, the computational and data requirements for cutting-edge AI remain immense. Training and running large models are resource-intensive, creating significant barriers to entry and widespread adoption, especially for smaller organizations. While techniques like dual GPU fine-tuning and compact models like Gemma are promising, they address specific bottlenecks rather than the overall economic viability for mass deployment.

The reality of AI involves massive datasets and extensive computational power. Even when AI can be run locally, as with Ollama on Apple Silicon, the performance and capabilities are often constrained compared to cloud-based behemoths. This dichotomy—the need for immense resources versus the desire for accessible, widespread application—is a fundamental tension that may limit AI's ability to spark a new, broad digital wave. The pursuit of efficiency in AI, including strategies to Slash AI Costs with Primitive Talk, highlights these inherent constraints.

Ethical and Safety Hurdles

Beyond technical and economic limitations, AI faces significant ethical and safety hurdles. Debates around AI liability, as highlighted by OpenAI Backs Illinois Bill to Limit AI Liability, and concerns about AI's potential for misuse, echoed in discussions about AI Safety, Multilingual LLMs, and Guardrails, continue to shape the development landscape.

The potential for AI to be used maliciously or to produce harmful content necessitates robust safety measures and regulatory frameworks. Companies are actively working on solutions, such as AI guardrails and adversarial training, but these add complexity and cost. Ultimately, the slow pace of addressing these deep-seated ethical concerns can impede the rapid, transformative deployment needed for a new digital wave. The very caution that led OpenAI to withhold GPT-2 in 2019, though perhaps overblown in its specifics, points to an ongoing, fundamental challenge in aligning powerful AI capabilities with societal safety and ethical norms.

The Promise and Practice of Agentic AI

The concept of AI agents—autonomous systems capable of performing tasks—is often cited as proof of AI's revolutionary potential. Platforms like Enso aim to democratize agent deployment, and companies are actively developing agentic capabilities. Yet, even advanced agent functionality, such as those explored in AI Agents: Your Next Revolution?, primarily focuses on automating existing workflows or enhancing current digital experiences.

Linear's core AI Agent functionality, for example, focuses on automating issue creation from Slack messages. While convenient, this is an optimization of communication and task management, not a paradigm shift. Similarly, Snowflake's Cortex Agents are designed to provide data insights and automate workflows within its data cloud. These are powerful enhancements within their respective domains, but they don't represent a new substrate for digital interaction like the internet or mobile apps did.

Beyond Automation: Where's the True Disruption?

The disruption promised by true AI agents would likely involve autonomous decision-making across complex, unstructured environments, or the creation of novel forms of digital interaction. While progress is being made—evident in research and specialized applications—the broad market adoption of such transformative agents is yet to materialize. Areas like 'artificial societies' for simulating opinions are fascinating but remain largely academic.

The launch of Ilya Sutskever's SSI Inc. with significant funding suggests a belief in AI's potential for profound advancement, potentially hinting at future breakthroughs. However, the current market, saturated with agent features that automate routine tasks, suggests that we are still in the 'augmentation' phase, not the 'replacement' or 'revolution' phase that would define a new digital wave. The current wave of AI agents is more about doing existing jobs for us, rather than enabling us to do entirely new things.

Optimizing the Digital Wave

The evidence suggests that AI is currently acting as an optimizer for the existing digital wave, rather than the engine of a new one. Companies are leveraging AI to enhance productivity, streamline operations, and extract more value from existing digital infrastructure. This phase is characterized by refinement and efficiency gains, not by the creation of entirely new digital economies or user experiences.

The focus on features like Snowflake’s AI_COMPLETE function and Linear’s automated triage points towards a future where AI tools are deeply embedded into existing software stacks, making them more capable and user-friendly. This is a significant technological achievement, but it represents an evolution of the current digital paradigm, much like the advancements in operating systems and software applications that occurred during the PC era, rather than a distinct new wave.

When Will the Next Wave Truly Arrive?

Predicting the arrival of the next true digital wave is inherently difficult. It will likely require a convergence of technological breakthroughs, affordable infrastructure, and compelling new use cases that fundamentally alter human behavior and interaction with technology. Whether AI, in its current trajectory, can achieve this remains uncertain. It's possible that AI will be a key component within the next wave, rather than the wave-maker itself.

The historical precedent suggests that true waves are disruptive and redefine industries. If AI continues to manifest primarily as incremental enhancements and automation of existing tasks, it may signify that the current digital wave, characterized by pervasive connectivity and data, is nearing its zenith. The focus then shifts to maximizing the utility of our existing digital infrastructure, with AI playing a supporting, rather than leading, role.

Consolidation and Refinement

The evidence suggests that AI is currently more about refining and optimizing the digital wave we're already riding, rather than launching us into a new one. While AI capabilities are advancing impressively, their practical impact on broad productivity and new digital paradigms is far more muted than the hype suggests. Companies like Gusto, Snowflake, and Linear are integrating AI features, but these are largely augmentations of existing services.

This hands-on review indicates that AI enthusiasts expecting a radical, transformative shift akin to the internet or mobile revolution may need to temper their expectations. The era of AI as a standalone 'next big thing' might be giving way to a period of consolidation, where AI becomes an indispensable, yet integrated, component of existing and future technologies. It’s a powerful tool, but not a new dawn.

Recommendation: Invest in Applied AI

For businesses and individuals, the takeaway is to focus on the practical application of AI. Instead of chasing the abstract promise of a new AI wave, prioritize how current AI tools can enhance efficiency, improve existing workflows, and drive tangible results. This means looking at platforms that offer practical AI integrations, like those seen in Snowflake and Linear, and evaluating them based on their ability to solve real problems.

The current state of AI suggests a path of continued evolution within the existing digital framework. The truly revolutionary aspects of AI may still be on the horizon, but for now, the value lies in its application as a highly sophisticated tool for optimization and automation. Don't expect AI to be the rocket ship to the next digital frontier; think of it as the advanced engine making the current journey smoother, faster, and more efficient.

Platform	Pricing	Best For	Main Feature
Snowflake	Tiered, Contact Sales	Data Analytics & Cloud Data Warehousing	Cortex Agents & AI_COMPLETE function for data insights and workflow automation.
Linear	Free, Standard, Plus, Business	Project Management & Issue Tracking	Linear Agent for automated issue creation from Slack; Triage Intelligence.
Gusto	Core, Complete, Concierge	Payroll, Benefits, HR for Small Businesses	Incremental AI features focused on HR and payroll efficiency, not core AI transformation.
OpenAI	API: Usage-based; ChatGPT Plus: $20/month	Foundational AI Models & Research	Development of advanced AI models (e.g., GPT series); early caution on 'too dangerous' models.

Sources

Explore more insights on the future of AI and its practical applications on AgentCrunch.

Explore AgentCrunch