This article is from the WeChat public account: machine power (almosthuman2017) , the original title: “Is there a collective outbreak of the AI bubble next year? No, financing will reach a new high, but this may be the last carnival. “Picture from Visual China
An effective way to deal with a bubble is to wrap it with another larger bubble, which is often called a “nested bubble.” If you want to pursue sustainable prosperity, you must do two things at the same time: carefully eliminate the false inside and seriously expand the space outside.
As we on the artificial intelligence giant wheel, we are more and more frequently listening to people talking about the rapid growth of artificial intelligence, which leads to the formation of bubbles. Unicorns are increasing, but few who have grown out of the strange capital circle have grown up independently. number.
For the upcoming 2020, the future is not clear. Where is the giant ship going?
Almost all players involved in artificial intelligence have realized that AI is like an accelerator.
For example, the combination of AI and “Internet retro vent” RPA (Automation) has achieved the IPA (Smart Process Automation) . Comparing RPA to human hands and AI to human brain, IPA uses AI as the brain to command the smart hands of RPA to complete the work. However, AI is much more than a game of commanding the brain.
More importantly, AI has been integrated from a single application to all aspects of the enterprise, reshaping its position in the entire industry chain.
Recently, IDC and Forrester, two of the world’s leading market research and consulting agencies, released their forecasts for the industrialization of artificial intelligence in 2020 and beyond. By IDC and Forrester ’s investigations and predictions, the future we can see is:
1) Whether there are differences in high-skilled engineers will create a “digital divide” between different companies.
2) Three major public relations disasters will make AI “infamous”: the widespread spread of Deepfake, the improper use of facial recognition technology, and uncontrolled personalization.
3) 25% of Fortune 500 companies will invest in IPA projects with “significant efficiency gains”. Nearly half of software and services companies will emphasize IPA in their portfolios.
4) “Data reliability” (digital trustworthiness) will become a key corporate asset. Forbes Global Listed Company 2000 More than 70% of the strong will monitor its “data reliability”.
5) 75% of enterprises will invest in employee retraining and development to meet the new skills needs and working methods brought by the application of artificial intelligence.
6) Even four out of every five AI conversations fail the Turing test. But by the end of 2020, conversational AI will still successfully support one-fifth of customer service interactions.
7) By 2023, nearly 20% of servers optimized with AI or coprocessor optimization will be deployed at the edge.
8) By 2025, 50% of computer vision and speech recognition models will run on edge (including terminal) .
9) At least 90% of new enterprise apps will include embedded artificial intelligence capabilities. However, APPs dominated by disruptive artificial intelligence will only account for 10% of the total.
10) AI financing in 2020 may reach a new high, but this will be the last crazyHuan——The artificial intelligence startup ecosystem will reach saturation.
Opportunity and risk coexist, but AI will not be late
Forrester said that while typical “capital incidents” may make companies wary of artificial intelligence, “brave companies” will continue to invest in AI and take an important step.
According to Forrester investigation:
53% of global data and analytics decision makers say they have implemented, are implementing, or are expanding or upgrading artificial intelligence.
29% of global developers (manager level or higher) have worked with artificial intelligence / machine learning software in the past year.
54% of global policymakers who use edge computing say that one of the biggest benefits they expect from edge computing is the flexibility to address current and future AI computing needs.
16% of global B2C marketing decision makers plan to increase spending by 10% in 2020 on data and analytics technologies, including artificial intelligence.
Forrester predicts that by 2020, 25% of Fortune 500 companies will automate their robotic processes. (RPA, Robotic Process Automation ) Add artificial intelligence building blocks such as text analysis and machine learning to create hundreds of new intelligent process automations (IPA, Intelligent Process Automation) use case.
“Automation requires artificial intelligence, and artificial intelligence also needs automation to scale,” Forrester said.
25% of Fortune 500 companies are shifting AI investments to more mundane, shorter-term, and moreTactical IPA projects with “very significant efficiency gains”. Nearly half of AI platform providers, global system integrators and managed service providers will emphasize IPA in their portfolios.
Based on the successful experience of these IPA use cases, IDC predicts that by 2022, 75% of enterprises will embed IPA in technology and process development, and use artificial intelligence-based software to calculate rules of operation and experience to guide innovation .
By 2024, artificial intelligence will be integrated into every part of the enterprise, and 25% of artificial intelligence solutions will be invested as “results as a service” to promote scale innovation and superior business value.
Artificial intelligence will redefine the user experience and establish a new user interface. Over 50% of user operations will be enhanced by technologies such as computer vision, speech, natural language, and AR / VR. In the coming years, we will see emerging user interfaces such as artificial intelligence and computer vision, natural language processing, and gestures embedded in every product and device.
However, emerging technologies are also double-edged swords, and the other side of their widespread popularity is also high risk.
Forrester warns that by 2020, the abuse and potential harm of artificial intelligence will multiply, and the public relations disasters in three areas will be noticed, making AI “infamous”: the widespread spread of Deepfake, Fair use, and unrestrained personalization.
IDC predicts that by 2021, 15% of customer experience applications will continue to be ultra-personalized by combining various data and updated reinforcement learning algorithms. Nonetheless, Forrester emphasised the positive side, and he still believes that “these measures will not delay the implementation of artificial intelligence plans for next year.”
Rather, they will emphasize the importance of designing, testing, and deploying responsible artificial intelligence systems—thinking of bias, fairness, transparency, interpretability, and accountability for sound AI governance.
The new digital divide
IDC predicts that by 2022, “data reliability” may be due to several high-profile public relations disasters (digital trustworthiness) will become a key corporate asset, with more than 70% of Forbes Global 2000 listed companies going through a formal plan to monitor their “data reliability.”
Forrester says leadership is important, and companies with chief data officer (CDO) use artificial intelligence, machine learning, or deep learning The probability is 1.5 times that of a company without a CDO. By 2020, senior managers such as Chief Data and Analytics Officer who takes AI seriously (CDAO) and CIOs will see that the data science team owns them Required data.
Forrester said that the real problem is “getting data from a complex application portfolio and getting approval from the heads of the various data departments.”
IDC believes that “The effective use of intelligent automation requires a lot of effort, especially in data cleansing, integration, and management, and the strong support of the IT department. For large enterprises, it is important to solve “It is a very big challenge.”
The application of artificial intelligence in all companies is not consistent. We see a new digital divide, that is, a divide between companies with or without highly skilled engineers.
Labor force is being redefined
Forrester believes that by 2020, “technical elites” will improve artificial intelligence and design skills, while others will be “clumsy.” Human-centered design skills and artificial intelligence development capabilities will be key.
IDC predicts that by 2024, 75% of companies will invest in employee retraining and development, including third-party services, to meet new skills needs and ways of working with artificial intelligence applications.
The composition and definition of “labor” are constantly being enriched.
IDC predicts that as intelligent automation expands across the enterprise, IT organizations will increasingly manage and support IPA support. ( Intelligent Process Automation (AI + RPA). Another major increase in labor will be the army of chatbots, assisting businesses with various tasks.
Forrester predicts that four out of every five AI conversations will fail the Turing test.
Despite this, by the end of 2020, conversational AI will successfully support one-fifth of customer service interactions.
The area of work affected by AI will also continue to expand.
IDC says that as computing power moves from the data center to the edge, managing and controlling edge processing equipment will face challenges.
By 2023, nearly 20% of servers optimized with AI or coprocessor optimization will be deployed at the edge. By 2025, 50% of computer vision and speech recognition models will run on edge (including terminal) .
Precursors of a bubble burst
AI will be everywhere.
IDC estimates that by 2025, at least 90% of new enterprise apps will include embedded artificial intelligence capabilities. However, IDC added that apps that are truly dominated by disruptive artificial intelligence will only account for 10% of the total.
So we need to wait another five years to see the “destructive” potential of AI explode.
Another Forrester forecast report states, “By 2020, as expectations return to reality, the prosperity of artificial intelligence will gradually increase.”
Forrester predicts that AI financing will reach a new high by 2020, but this will be the last carnival-“There are more than 2,600 companies worldwide, and the artificial intelligence startup ecosystem has reached market saturation.” Forrester said that the most important signal of the future economic slowdown is that in the past 12 months, 20 artificial intelligence companies have raised round after round of financing, which is comparable in size to unicorns.
“This is not sustainable.” Forrester said.
This reminds me of Charles McKay (Charles Mackay) in “Extraordinary mass delusion and crowd madness” Described: “The bubble was then full-blown and began to quiver and shake preparatory to its bursting.”
(“The bubble is bursting out in full and is beginning to tremble in preparation for bursting”)
This article is from the WeChat public account: machine power (almosthuman2017)