Artificial intelligence ownsBroad commitmentDifferent industries and sectors. The most important artificial intelligence communicator in the worldAndrew Wu(Andrew Ng)Listed the creation of an artificial intelligence strategy as itsAI conversion manualThe key element. As artificial intelligence will change every industry, how do you create a sensible strategy to harness its power? What constitutes an artificial intelligence strategy? What is the difference between a startup and a company's artificial intelligence strategy?
Artificial intelligence strategy modeling is different from creating traditional business strategies. This article is intended to guide practitioners in developing targeted artificial intelligence strategies. These insights are based on my personal experience andVolkswagen.Google.Artificial intelligence fundInterview with experts from the company's global artificial intelligence leaders. At the end of this article, you'll learn how artificial intelligence strategies relate to business strategy, the core components of an artificial intelligence strategy, and how to separate good and bad artificial intelligence strategies.
Let's start planning. 🤔
Business strategy and artificial intelligence strategy
The artificial intelligence strategy exists to support business strategies. Business strategy defines the way forward for your company. Which initiatives will provide the highest business value? What features separate the company from its competitors? Express business strategies through measurable goals, such as with key performance indicators (KPI) or goals and key outcomes (OKR)form. These methods track the progress of achieving business goals.
German China Artificial Intelligence AssociationVice chairmanRaphael Kohler saidIt is crucial to have a good understanding of the goals of the artificial intelligence strategy. He emphasized that discussions cannot be technology-centric and must be driven by business value.
There is no one-size-fits-all strategy for artificial intelligence and business strategy. - Dominik Haitz
Business coordination to achieve these KPIs. The work of the artificial intelligence strategy is to provide a way to influence the key performance indicators of these businesses.
The opportunities and challenges of artificial intelligence
Balancing AI's capabilities and limitations is key to understanding how AI supports your business strategy.
In general, AI can do wellthree things: Automate processes, create new products or improve existing ones. An example of an automated process isRobotic process automationIt frees employees from cumbersome, repetitive tasks. AI can also help companies generate new products, such asSmart home speakerPowered by AI. Finally, AI can also improve existing products. Most of today's credit decisions are supported by artificial intelligence. This model has more information than humans, which reduces it.consumerOfOverall credit cost.
However, AI struggles in complex tasks. These tasks may include contextual understanding or any need to exceed1 people's things. AI is very good at performing tasks defined in a narrow sense and does not perform well in an uncertain environment.
Knowing the scope and boundaries of AI is critical before planning an AI job.
Now that you understand the relationship between artificial intelligence strategy and business strategy and what opportunities and challenges exist, how do you create one?
The core component of the artificial intelligence strategy
Just as every company is affected by electricity, every company is affected by artificial intelligence. Although no AI strategy looks the same, all AI Strategies need to answer similar questions. The core components of any artificial intelligence strategy involve the trinity of data, infrastructure, and algorithms, surrounded by the pillars of skills and organization. Let's delve into each component.
Without data, there is no artificial intelligence. The data relates to all information related to improving the business. It can be anything from sensor data from autonomous vehicles to financial data for business decisions. Creating a data strategy is an important part of any AI strategy. Andrew Ng recommends touching the following:
- What data can you get strategically?
- Do you collect all or selected data?
Artificial intelligence fund AI investorJason RischEmphasize the importance of the right timing of strategic data collection. Flooding in the data should no longer be an option. Jason witnessed " select* “Data strategy has failed in both startups and companies. Startups that focus on creating models before building viable products waste valuable resources. Similarly, companies that get startups because of too much data often cannot find value in them. This is a common fallacy in the healthcare industry, and companies want algorithms to find patterns in random data. The key is to collect *right* data.
In the past decade, data has been the main source of corporate growth. Data-driven decision making is the key to success, so you need to define a powerful data strategy. - Tari Singh
Now that you've created your data strategy, it's time to consider the next infrastructure.
The second core component of the artificial intelligence strategy is the infrastructure. Infrastructure involves making data accessible and providing the computing power needed to process data. The AI model is eager to acquire computing power, and your AI team needs an infrastructure to develop and deploy models. Ideally, this infrastructure can be adapted to your company's needs.
A unified data warehouse centrally accesses data available throughout the company. In traditional companies, you will find data that is hoarded in silos that other teams cannot access. This usually has structural, organizational and legal reasons. However, establishing contact across business team-specific data is at the heart of artificial intelligence efforts. Data scientists are good at finding patterns in data, so your goal is to give them access to as much data as possible.
The important issue is if you rely on cloud services or build your own infrastructure for AI.AWS.MicrosoftOrGoogleWait for cloud providers to provide out-of-the-box AI solutions. When using the cloud, you only pay for the cost you spend. There are also plenty of resources to help you set up quickly. You also continually invest time and resources to maintain the server. When you buy your own hardware, the upfront cost is higher. Cloud solutions start at a lower cost, but in the long run, invest in their ownInfrastructureGet a return. The strengths and weaknesses depend on your industry, so it's important to determine your needs before making a decision.Deepkapha.aiCEO and co-founder Tarry SinghIt is not recommended to focus solely on the cloud as a company that develops algorithms as a competitive advantage.
Once you know how to use hardware for AI, consider the algorithm part of the next AI.
Algorithms are the pinnacle of the AI Trinity because they use data and infrastructure to produce valuable products. The algorithmic part of the artificial intelligence strategy is tricky. Answering these questions will take you a step further.
- Is proprietary algorithms a key driver of business value?
- Are you an open source model or do you prefer to retain their exclusive rights?
Google Chief Decision ScientistCassie KozyrkovThink existTwo machines learning the world. Cassie likes to distinguish between machine learning research and applied machine learning. Conducting research requires a different approach than applying existing algorithms.
The artificial intelligence community has become better at publishing reusable public datasets and models. This gives your company a huge advantage because you have access to a variety ofAI Model Zoo. The main problem you should answer in the AI strategy is that the algorithm is the primary business driver for AI functionality. If so, you should set up a patent plan and motivate employees to apply for a patent. If not, you should consider open source models and use crowdsourcing knowledge to improve the algorithm.
Next, let's take a look at the skills needed to use AI in a company.
Once the trinity of artificial intelligence is in place, you need people to achieve their own destiny. The core of people is to use data, infrastructure and algorithms to generate business value. How do you authorize people in your organization to use AI? Answer the following questions in your AI strategy:
- Are you building an internal team or an outsourcing mission?
- How do you continually educate management and employees about AI?
Andrew Ng suggested building an internal AI team. AI provides domain knowledge and is difficult to outsource in certain industries. External consultants may not know your data, infrastructure and issues, and your own employees. Therefore, a viable approach is to bundle enthusiastic employees and educate them about artificial intelligence.
1&1 IonosData scientistDominik HaitzSaid that AI as a new technology is different from other technological innovations. People are often not only unaware of the actual capabilities of artificial intelligence, but they are often misunderstood. This can range from “all-powerful threats to humanity” to the concept of a multi-functional system that works out of the box.
Once the internal team is in place, they need to act as facilitators. The promise of artificial intelligence is too large to be packaged in a single team. The artificial intelligence strategy should implement a program that constantly educates everyone to find artificial intelligence use cases. Often, these plans should target high-impact individuals who can invest in AI projects.peopleCorrectTodoku.aiCo-founderRachel Berryman(Rachel BerrymanI am convinced that artificial intelligence is essential to the understanding of managers because it will gradually penetrate AI as an opportunity for employees in opportunities.
Let's study the last component of artificial intelligence strategy-organization.
The last but arguably the most important component of an artificial intelligence strategy is to prepare artificial intelligence for your organization. Specifically evaluate your organization's design and development process. Then, align them with best practices.
- How do you make your AI team deliver business value across teams and domains?
- Is your process ready for the ML workflow?
The benefits of artificial intelligence are omnipotent. The most important thing is to understand that AI can't work in an island. AI is not working in a vertical, customer-centric business unit, but can be seen as a horizontal driver for the company. AI can influence internal processes, create new products, or improve existing products. To this end, Andrew Ng suggested establishing an independent unit that became the core enabling point for the entire company's AI. The department then works with existing departments to find high-impact AI projects and support their implementation.
To enable AI throughout the company needs to understandMachine learning workflow. Machine learning follows a highly iterative process and the results are far from certain. you can use itAI Project CanvasTools such as to assess the likelihood of success, but you can hardly guarantee specific results. The exploratory nature of AI makes it difficult to track target measurements across the company.
If the data is not thoroughly evaluated, the working model cannot be guaranteed. So if you don’t invest firstETLAnd initial data analysis makes it difficult to estimate the specific business impact of an AI project. - Rachel Berriman
Consider your process: Are they ready to support AI? If you work in a safety-critical industry, there may be no process for validating the statistical learning model. Does your company follow the waterfall engineering process? Rethink the current development process and check if it is consistent with the machine learning workflow.
Now that you understand the core components of the AI strategy, let's take a look at more tips for avoiding common pitfalls.
Good, bad and ugly artificial intelligence strategy
Who do you need to develop an artificial intelligence strategy in your team? What constitutes a good or bad AI strategy? What is the difference between an enterprise and a startup in an artificial intelligence strategy? The final summary is intended to answer previous questions.
AI Strategy Team
Developing an artificial intelligence strategy is an effort of the team. You need different perspectives across the core components of the AI strategy. The team mix between startups and companies is different. Startups create artificial intelligence strategies in smaller teams, around the technical feedback of data engineers and business feedback from product owners or business developers. The corporate team involves more features. Create an internal artificial intelligence strategy for the world's largest automakerAndreas MeyerKnow that in a company with a professional role, you need a lot of domain knowledge to find a viable artificial intelligence strategy. You need to have a large group of people with different roles in the company, and in startups, you can create great AI strategies with a few generalists.
Good and bad artificial intelligence strategy logo
Good and bad AI Strategies have common characteristics in each company. A good artificial intelligence strategy is affected, well supported by the company, and adequately funded in terms of time, salary and expectations. The bad AI strategy is hype-driven, focusing on technology rather than impact, and hiring 2-3 data scientists to compete for projects. Try to deviate from the latter.
Corporate and startup AI strategy
Creating an AI strategy is different for companies and startups. Raphael Kohler explains that companies must consider legacy systems and challenge existing organizations' change management, while start-ups can focus on人工智能OfVirtuous cycle. Andreas Meyer knows that the path to creating influence for artificial intelligence can be overwhelming. He said that in large enterprises, there is a lot of potential to use artificial intelligence to automate processes. For Andreas, it is important to start and provide value.
On the other hand, startups should focus on providing a product that works well without artificial intelligence but still steadily improves the customer's use of the product. Then analyze customer interactions to improve the product to attract more users. Once they enter the virtuous circle of artificial intelligence, artificial intelligence startups will embark on the road to success.
Data is the oil that powers the AI engine and cannot underestimate the importance of iteratively improving the product by getting the initial setup and the right information from the customer. - Jason Risch
The core components of the artificial intelligence strategy are intertwined and interdependent. Core components may have different importance in different industries, but they are always relevant.
- Artificial intelligence strategy should always serve a higher corporate strategy
- The core components of an artificial intelligence strategy are data, infrastructure, algorithms, skills and organization.
- The AI strategy team should be composed of product managers, data scientists and business developers.
- A good artificial intelligence strategy focuses on medium-term goals and a holistic approach, rather than the hype of some data scientists driving employment
- The right time for strategic data collection can determine your AI fate
In this article, you learned how to create an AI strategy. When creating an AI strategy, consider the core components of AI. We are looking forward to a hugArtificial intelligence implemented for ten yearsworld.
This article was transferred from awardsdatascience,Original address
My blog will be synced to Tencent Cloud + Community and invite everyone to join us:https://cloud.tencent.com/developer/support-plan?invite_code=ksay66dn578f