"One reason why people are so interested in the diffusion of innovations is that even when a new idea has clear advantages, it is often difficult to adopt," Everett Rogers said in his seminal 1983 publication He was ostensibly a pioneer on the subject in Diffusions of Innovation, 3rd ed., 1983). As Dr. Rogers points out, the idea was not original to him; has always been part of the human condition. As Niccolò Machiavelli wrote in his 1513 letter (which became his classic The Prince), he observed 470 years ago:

"There is nothing so difficult to plan, so doubtful to succeed, and so dangerous to manage, than to create a new order... Whenever his enemies have an opportunity to attack the innovator, they attack him with guerrilla zeal, while others defend him idly, so that both the innovator and his party are vulnerable" (Machiavelli, 1532).

Thereafter, the British Navy in 1747, the American inventor and founding father Ben Franklin in 1781, the French judge and lay scientist Gabriel Tard in 1903, the British anthropologists Edward Gifford and Alfredo The adoption or spread of innovations was strongly mourned by de Krober in 1937, researchers Bryce Ryan and Neil Gross in 1943, and at least in 1953 between 1941 and 1962 For 21 years, the authors published papers in peer-reviewed journals (Rogers, 1983) (see Figure 1).

Dr. Rogers defines "diffusion" as:

“…the process of innovative communication between members of a social system over a period of time through specific channels; it is a special type of communication because information is related to new ideas” (Rogers, 1983).

This novelty inherently involves uncertainty. In this case, uncertainty includes perceptions of alternatives to new ideas, and comparative probabilities of the effectiveness of those alternatives, including the status quo (Rogers, 1983). In modern times, many innovations are technologies, which Rogers goes on to define as: “the design of an instrumental action that reduces the uncertainty in the causality involved in achieving a desired result” (Rogers, 1983). Technological innovations thus create uncertainty in potential adopters' perceptions of their efficacy relative to alternatives, while at the same time providing opportunities to reduce uncertainty by applying faster and more accurate causal linkages (Rogers, 1983). One could reasonably argue that the diffusion of technology is the second derivative of uncertainty, that is, the uncertainty in potential users' perceptions of whether the technology will reduce uncertainty.

Modern innovation diffusion theory was originally based on the adoption of new methods in agriculture and home economics in the 1950s, and Rogers extended the theory to technologies involving hardware and software beginning in the 1960s (BEAL, 1957). It is through this process that these perceived uncertainties in the minds of adopters are magnified or reduced, as the approach, culture, nature of adopters and their areas of focus determine the speed of innovation diffusion or technology adoption. These factors are embodied in the policies governing an organization, determining when and when an organization's raison d'être—from the military to manufacturers to health care—provides new capabilities.

roger bell curve

Rogers hypothesizes that, under the theory of diffusion of innovation, the rate at which technologies are adopted can be depicted as a normalized Gaussian distribution on the x-y axis—or a “bell curve”—first familiar with a Cartesian coordinate system. In it, Rogers found and showed that adopters were divided into five segments depending on their descent in this adoption chronology. The earliest adopters are the "innovators", accounting for 2.5% of the market. The second adopter in chronological order is the "early adopters", accounting for 13.5% of the market. "Early majority" adopters came in third, accounting for 34.5 percent of the market. "Late majority" adopters accounted for 34.5% of the market, ranking fourth. Moreover, the “laggards” accounted for the bottom 16% of the market (Rogers, 2003) (see Figure 2). Furthermore, Rogers hypothesizes that each class of adopters goes through four cognitive stages: (1) awareness; (2) decision to adopt or reject; (3) initial use; (4) continued use; Among them, the five factors that most influence adopters are: (i) relative advantage; (ii) compatibility; (iii) complexity; (iv) testability; (v) observability (LaMorte, 2018).

While this level of understanding of the steps in the technology adoption process and their causal relationships has been successful across many disciplines, it also contains elements that can be a deficit in healthcare and public health (LaMorte, 2018). Specifically, because the model originated outside of healthcare and public health, it: (a) did not include the participatory approach typically required in healthcare to ensure acceptance of the "six Ps": patient, provider, payers, drug manufacturers, suppliers, and policy makers; (b) is more applicable to behavioral adoption rather than behavioral cessation, which is a major concern because in modern medical technology most innovations are replacing existing technologies and, (c) it does not take into account the resources, social and peer support of organizations or adopters in adopting new technologies (LaMorte, 2018).

Evolution Roger Bell Curve

While all academic and conceptual theories are constantly tweaked and tweaked by new assumptions and discoveries, between 1962 and 2015 there were five major evolutions in the bell curve that Rogers touts as a model for innovation diffusion and technology adoption. The first major evolution related to this was the technological S-curve, initiated by Richard Foster in 1986 and applied more broadly by Clayton Christensen in 1997 in his seminal book The Innovator's Dilemma (Foster, 1986) ( Christensen, 1997). Foster reasoned that technological innovation can be represented by cost and/or time on the x-axis and technological performance progress on the y-axis, where the curve or line for a new technology is always some form of "S", the induction of a new technology Time ("R&D") is fundamental, return through adoption or return on investment is vertical, and market saturation and obsolescence are the top of the "S" (Foster, 1986). Second, Christensen pointed out that these “S” curves are accompanied by a series of fluctuations (see Figure 3), in which the key factors determining their success are: (a) the time frame in which organizations enter the curve to avoid being overwhelmed by more prescient competitors elimination; and, (b) their ability to continue to innovate without interruption to keep these “S” waves going over the long term (Christensen, 1997). Third, Christensen goes on to identify two key reasons for whether technological innovations are adopted or not, and the speed of adoption or rejection; this has to do with the relative needs and resources of adopters. If the status quo satisfies the needs of potential adopters within available resources, they will cling to the status quo and delay or reject innovation. Likewise, if a technological innovation is not within the adopter's given resources, the adopter will cling to the status quo and reject or delay the innovation regardless of whether there is a perceived need in some cases. These reasons are key to the adoption of AI and similar new technologies in healthcare organizations (Christensen, 2015).

The next major evolution in the proliferation of innovation-based technology adoption lifecycles is about gaps or rifts, which occur on the third chronology but are listed as the fourth year for the sake of coherence and clarity. surface. In 1991, Geoffrey Moore pointed out in his book "Crossing the Chasm" that a large number of technological innovations pass through the induction/research and development stage, are welcomed and used by early adopters, but due to For various reasons, it has never been more widely adopted by the market (see Figure 4) (Moore, 1991).

Oversimplifying, Moore argues, there is a divide between early adopters and majority adopters because they have substantially different psychological profiles as to how and why they make decisions. Innovators and early adopters have a pro-adoption bias because they have an intrinsic appreciation for new capabilities; they tend to like, want, and adopt them. However, 68% of the market makes up its early and late majority, which is more focused on utility—the kind of needs and resources Christensen writes about. Most market players are also skeptical, often based on experience, knowing that the vast majority of new technology innovations never go far or last (Moore, 1991). Later majority adopters have the same proportions as early majority adopters, and they differ in that they lack confidence in their ability to implement organizational change (Moore, 1991). To overcome these skeptics, Moore argues, distinguishing a technology's existence requires a great deal of education, marketing, and relationship building, which in turn requires staying power, which in turn requires capital — more than most companies have or can raise. Capital is even greater, creating a "valley of death" for technological innovation start-ups (see Figure 5) (Moore, 1991).

Fifth, and finally, from 1998 to 2008, Carl May and his colleagues proposed Normalizing Process Theory (NPT) to develop previous models that helped explain innovation diffusion and technology adoption lifecycles in healthcare (2009 in May). A Theory of Normative Processes addresses three core issues related to the adoption of technology in health care settings: (1) Implementation—the social process by which new actions are implemented; (2) Embedding—the incorporation of these new practices into habits and routines;( 3) Integration - the process of replicating and sustaining new practices across the organization (May 2009). The Theory of the Normalizing Process states that: (A) practices become ingrained and routine as a result of the concerted efforts of individuals to act; operate to facilitate or inhibit the mechanisms through which human behavior is expressed”; (C) replicating practices organization-wide requires sustained support and investment from the change agent collective within the organization (May 2009).

The AI Glacier in Healthcare

It is a fact that artificial intelligence will disrupt healthcare from reactive to predictive and proactive, and extend our lives by decades through personalized medicine, if one works with experienced and publicized The widespread or timely adoption of these tools to deliver on this promise remains largely overblown, digital health entrepreneurs talk. Although an estimated $12 billion was privately invested in digital health companies in 2017, many of them related to artificial intelligence, few have achieved blockbuster success that could justify private equity investment (Yock, 2018).

There seem to be five strands to the explanation for why AI adoption in healthcare has been so lukewarm, despite the remarkably good results of AI in a field that is vital to humanity.

First, the explanations given by health technologists focus on the idea that most digital health and AI startups follow the wrong model, one that has succeeded with consumers and products in other industries but ignored fundamental difference in healthcare (Yock, 2018). Tested and proven technology launch strategies in other industries focus on bringing a minimum viable product to market quickly, and then iterating on new versions and releases based on feature and function sets that have proven successful among early end users (Yock, 2018). This strategy ostensibly ignores stakeholder complexity, risk aversion, and the regulatory environment of the healthcare industry (Yock, 2018).
Second, the “valley of death” that Moore describes is deepening in the medical field as adoption cycles lengthen. Startups must survive longer and do more marketing and prospect education, which requires more capital, to successfully overcome the additional hurdles Yock describes. In addition, technologists in the frontiers of data science, artificial intelligence, and start-ups are also in high demand in other industries, such as financial technology or consumer goods. As a result, it is more costly to retain this highly competitive talent for many years while healthcare slowly adopts new technologies.
Third, we must re-examine the diffusion of innovation theory, as it seems that health care-focused technologists have become overly reliant on its reductionist evolution, ignoring its initial warnings. First, we can look at the elements of innovation defined by Rogers: (1) Relative advantage; (2) Compatibility; (3) Complexity; (4) Testability (Rogers, 1983). In each of these areas, artificial intelligence is problematic in healthcare. AI is often incompatible with existing systems, policies, and processes, and they need to be replaced. Furthermore, AI is notoriously complex, beyond the knowledge of many users, and sometimes beyond comprehension; thus, they hate to accept what they cannot trust, and they cannot trust what they cannot understand. In addition, experiments with AI on many healthcare issues are troublesome because they touch on a critical area that affects human well-being, which is high-stakes and involves many ethical issues.

Second, as Rogers points out, change agents who initiate innovations and those who must gain social approval tend to be heterophilic—meaning they are held together by similarity in groups, each of which is related to Other groups are very different. Consequently, change agents are often more technologically advanced than users, and thus are not conducive to effective mutual understanding in communication.
Third, Rogers argues in his book The Importance of Scientific Validation of Innovation (Rogers, 1983), which is probably the most practical and influential work for AI startups in the medical field. In healthcare, this means clinical trials; however, for software clinical trials, there are few, if any, as widely accepted standards as drug trials. Also, it's worth mentioning that most trials are very expensive and the academic medical institutions that are able to do them (Massachusetts General Hospital, etc.) see it as a way to use the "stamp of approval" to generate extra revenue, all of which Both increase the depth and breadth of Moore's "Valley of Death". In short, there is a shortage of funds to pay for software trials of artificial intelligence applications in healthcare. As a result, the vast majority of innovations are never scientifically proven, and many of the innovations that would have been adopted early have been shown to be scientifically flawed, to the extent that the skepticism and chasm before most adoption widens — and ensues It is the "valley of death" in the technology adoption life cycle.
Fourth, recall Christensen's S-curve waves (Figure 4). Even if AI startups in healthcare can overcome these significant hurdles, this is just an adoption cycle, or the first S-curve in a wave of S-curves if they are to sustain themselves. One consequence is that a significant portion of companies in the medical field that are competitively challenged by AI must iterate through new innovations, or they risk becoming a shorter-lived single-product company (aka "one pony"). ").
Finally, we are fascinated by timescales, whether AI in healthcare is truly disruptive, and what disruptive really means. Christensen pointed out that the life cycle of technology maturity in the market is usually 15-20 years (Brown, 2006). This period is not really a disruptive one, but a transformative one. If we study other transformative technologies such as e-mail and the Internet, we verify Christensen anecdotally because decades have passed from their invention to widespread use. For AI start-ups in healthcare, which already face an additional long and deep chasm in the technology adoption life cycle, and the competing need to innovate constantly, the duration of this transition greatly amplifies the impact of each innovation. "Death Valley".

The key to solving the many challenges facing AI startups in healthcare may lie in Christensen's correct definition of disruption. Christensen believes that it is misleading when it disrupts the capabilities of a new technology applied in a product or service (Christensen, 2015). Instead, according to Christensen, disruption is a process—disruptors start with small-scale experiments at the low end or at the edge of the market (“the fringes”), and focus on how needs change and evolve over time, leading to new business models (Christensen, 2015). Disruptors find a new paradigm to meet nascent and evolving customer needs and completely replace or replace one technology with another, often taking decades (Christensen, 2015). Start-ups, however, are freed from many of the competitive pressures of continuous innovation because they are not seen as a center of competition or a threat, are less costly, can survive Early customers provide feedback that can address many of the structural idiosyncrasies in the market for technology adoption.

This article: https://architect.pub/golden-ai-glacier-rethinking-rogers-bell-curve-healthcare
Discussion: Knowledge Planet [Chief Architect Circle] or add WeChat trumpet [ca_cto] or add QQ group [792862318]
No public	【jiagoushipro】【Super Architect】 Brilliant graphic and detailed explanation of architecture methodology, architecture practice, technical principles, and technical trends. We are waiting for you, please scan and pay attention.
WeChat trumpet	[ca_cea] Community of 50,000 people, discussing: enterprise architecture, cloud computing, big data, data science, Internet of Things, artificial intelligence, security, full-stack development, DevOps, digitalization.
QQ group	[285069459] In-depth exchange of enterprise architecture, business architecture, application architecture, data architecture, technical architecture, integration architecture, security architecture. And various emerging technologies such as big data, cloud computing, Internet of Things, artificial intelligence, etc. Join the QQ group to share valuable reports and dry goods.
video number	[Super Architect] Quickly understand the basic concepts, models, methods, and experiences related to architecture in 1 minute. 1 minute a day, the structure is familiar.
knowledge planet	[Chief Architect Circle] Ask big names, get in touch with them, or get private information sharing.
Himalayas	[Super Architect] Learn about the latest black technology information and architecture experience on the road or in the car.	[Intelligent moments, Mr. Architecture will talk to you about black technology]
knowledge planet	Meet more friends, workplace and technical chat.	Knowledge Planet【Workplace and Technology】
LinkedIn	Harry	https://www.linkedin.com/in/architect-harry/
LinkedIn group	LinkedIn Architecture Group	https://www.linkedin.com/groups/14209750/
Weibo‍‍	【Super Architect】	smart moment‍
Bilibili	【Super Architect】
Tik Tok	【cea_cio】Super Architect
quick worker	【cea_cio_cto】Super Architect
little red book	[cea_csa_cto] Super Architect
website	CIO (Chief Information Officer)	https://cio.ceo
website	CIOs, CTOs and CDOs	https://cioctocdo.com
website	Architect practical sharing	https://architect.pub
website	Programmer cloud development sharing	https://pgmr.cloud
website	Chief Architect Community	https://jiagoushi.pro
website	Application development and development platform	https://apaas.dev
website	Development Information Network	https://xinxi.dev
website	super architect	https://jiagou.dev
website	Enterprise technical training	https://peixun.dev
website	Programmer's Book	https://pgmr.pub
website	developer chat	https://blog.developer.chat
website	CPO Collection	https://cpo.work
website	chief security officer	https://cso.pub ‍
website	CIO cool	https://cio.cool
website	CDO information	https://cdo.fyi
website	CXO information	https://cxo.pub