Jack Dorsey has reopened a basic question about the technology industry’s most familiar term. In an August 6 post, Block’s chief executive wrote, “Artificial intelligence” is the worst descriptor. His brief remark moves attention from model performance to the language surrounding AI. It also raises questions about how names shape trust, adoption, and commercial expectations.
Dorsey’s wording follows Block’s effort to build an “intelligence layer” across its operations. He and partner Roelof Botha describe a company organized around shared models instead of management chains. Their framework places people at the operating edge, where they handle context, judgment, and difficult decisions. That description presents AI as infrastructure supporting work, rather than a separate artificial actor.
Why Artificiality Creates Unease
Merriam-Webster defines “artificial” as something humans make, often to resemble something natural. Its listed synonyms include fake, synthetic, unnatural, and simulated. Those meanings do not describe every use of modern AI, yet they accompany the term in daily language. By contrast, digital describes a format, while smart usually signals an added function or capability.
Intelligence is also a heavy word, as it typically refers to human thinking and comprehension. NIST employs less broad terminology for AI systems. It describes them as systems that use machines to make predictions, recommendations, or decisions for goals set by humans. That technical description makes objectives and outputs the focus but does not mean that software thinks like a person.
Several studies have been undertaken to validate the idea that labels affect evaluations of automated systems. There are some studies that corroborate the idea that labels do influence the assessment of an automated system.
The CHI study adopted the following terms: AI, algorithm, computer program, automated system, and statistical model. Language impacted perceived complexity, fairness, and trust in both studies (397 subjects and 622 subjects). One name could not be confirmed as a secure name in each context in the study, however.
Public caution creates a hard setting for any AI brand. Pew Research Center reported that half of U.S. adults felt more concerned than excited about wider AI use. Only 10% felt more excited, while 38% reported equal levels of both feelings. Those findings measure views of AI broadly and do not isolate the word “artificial.”
Old Tech Names Shaped Expectations
Meanwhile, earlier technology labels show how names can simplify complex systems. Cloud computing replaced descriptions of remote servers, shared storage, and internet-based processing with one compact metaphor. NIST later standardized cloud around on-demand access to shared computing resources. Its standard also lists rapid scaling and measured service among the defining features.
Big data followed a different route and made scale the central message. NIST describes it as the large volume of information produced across a networked and sensor-filled world. The agency later developed a common framework using the term across data architectures. However, the name describes scale and not the exact tool, method, or business result.
Cryptocurrency tied cryptography directly to money, giving unfamiliar software a familiar economic frame. The word joins its technical method with a financial function. Regulators later adopted broader terms, including crypto assets and digital assets, for legal and market documents. The SEC now defines a crypto asset through its cryptographically secured distributed ledger record.
Block Reframes AI as an Operating Layer
Notably, Dorsey’s preferred language already appears in Block’s new operating model. The company’s framework places capabilities, company data, customer data, and interfaces around an intelligence layer. That layer composes services when its models identify a specific customer need. It functions as coordinating software within the business, rather than one product sold under an AI label.
Block also organizes staff around individual contributors, directly responsible individuals, and player coaches. Its published framework assigns routine information flow and alignment work to the shared system. People handle specialist work, customer contact, cultural context, ethical choices, and high-stakes situations. Sequoia’s podcast description says Block cut 40% of its workforce during this broader restructuring.
Block’s wording separates the software layer from the worker roles surrounding it. “Artificial intelligence” can suggest a substitute for human intelligence, especially when companies discuss job cuts. “Intelligence layer” instead describes a central operating function connecting data, services, and employees. Still, a wording change does not alter workforce numbers, costs, controls, or system performance.
New Names Carry Different Meanings
However, several alternatives offer different frames, and each changes the expected meaning. Machine intelligence identifies the system’s origin, although it still compares software with human ability. Cognitive computing stresses reasoning and decision support. IBM uses that term for systems that simulate parts of human cognition.
Intelligent software sounds more familiar, yet it covers products with different capabilities and risks. A narrower product name may provide clearer information than one that replaces the entire field. Companies could use terms such as “forecasting model,” “writing assistant,” “recommendation engine,” or “fraud detector.” Those labels describe the system’s task before explaining its technical design.
Rebranding alone does not resolve underlying concerns about AI systems. Research shows that labeling can influence user trust, but its impact depends on context, user experience, and how the technology is applied. A naming change may shape perception, yet transparency, performance, and accountability remain the primary factors determining public confidence.
Regulators have also emphasized the need for accurate representation. The Federal Trade Commission has taken action against companies making unsupported claims about AI capabilities, reinforcing that terminology must align with actual functionality. Dorsey’s remarks highlight an ongoing debate over how the industry describes its tools, but any shift in language will need to balance clarity, accuracy, and public expectations.
Related: Michael Saylor Says Bitcoin Is Digital Capital and AI Is Digital Intelligence
Disclaimer: The information presented in this article is for informational and educational purposes only. The article does not constitute financial advice or advice of any kind. Coin Edition is not responsible for any losses incurred as a result of the utilization of content, products, or services mentioned. Readers are advised to exercise caution before taking any action related to the company.
