The hype around artificial intelligence (AI) and machine learning (ML) has reached incredible levels, with some commentators calling AI the fourth industrial revolution and others calling it new power.
I am one of the believers in this theory. A lot of money is being invested in companies in the AI ​​and ML sectors because they have the potential to revolutionize most industries, even if not all. Please see the picture below.
Avoid traps focusing on real problems
Such a major technological revolution should be widely and deeply financed, which may justify the huge investment in this field. But I am worried that many entrepreneurs are caught in the trap of focusing on the AI ​​infrastructure, algorithms and platforms rather than applications.
Successful companies usually start by solving a specific user problem and evolve over time until they finally build a broader "platform" based on a solution. The most important thing is to solve specific business problems, not the problems of the technology itself. However, considering the externalization of this technology, we are easily attracted by algorithms and models, ignoring these applications.
So, for the great AI and ML startups, what makes it great? Can be summarized as the following:
1. Eliminate or reduce human labor in areas previously considered difficult to automate; 2. Develop “white spaces†with new capabilities (previously not cost-effective new products and services); 3. By embedding ML technology into applications In the program, make traditional applications more valuable.
Why avoid the same level of ML platform?
Have you noticed that the machine learning platform does not seem to have a category? There are several reasons for this. Large network companies like Google and Facebook are not only investing in AI and ML, but also adopting open source tools and platform strategies. Given that these large network companies have access to vast resources and specialized unique data sets, it is difficult for start-ups to compete with them. If you want to differentiate your startup based on these capabilities, you will face a huge disadvantage compared to those big companies.
In addition, the widespread coverage of data science talent shortages has also affected customers' ability to leverage platforms and algorithms. The lack of artificial intelligence technology means that customers are not able to build their own AI and ML, so startups with the same level of platform will eventually become professional service roles that help each customer define and achieve their specific goals.
Another important aspect of entrepreneurs building platforms at the same level should take into account the complexity of the “marketization processâ€. Different vertical markets may have different buying behaviors, and you may need to solve different vertical problems through different channels. Of course, before you choose a vertical market, you should make sure that it guarantees sufficient scale and growth to support a large company.
Test industry
If your ML model can be applied to multiple industries, you need to consider the following variables before deciding which industry to focus on:
1. Deployment costs not only consider how much customers spend to buy your technology, but also the cost of turning their current solution into a new product. For example, if a Chinese manufacturing worker earns $6,000 a year and the cost of using a robot to replace the worker is $4,000, then there is at least a six-year minimum return period (excluding operating costs). This situation may not be very attractive to ordinary factory managers.
2. The extra value added by the extra value exceeds the cost, what value can your ML-based software provide? Can you provide better quality, increase customer satisfaction, reduce errors, improve performance or throughput? For example, in terms of recruitment, people have prejudice and preferences. As a result, startups like Gild, Entelo, and TexTIo have developed ML-based software that automatically hires employees without these biases.
3. Is there a lot of red tape for regulatory/compliance issues that may complicate your product adoption process? An obvious example is the self-driving car.
4. Target conflicts between potential customers AI and ML eliminate or reduce the scale of human labor, which may be much larger than any existing technology, and thus the resistance generated is much greater. Will your sales team lose your job because of your new technology? For example, one of the main concerns of IT outsourcing companies is that hourly fee-based maintenance is reduced due to "automation" impact.
5. Industry preparation Sometimes, due to extreme risk aversion, an industry is not ready to accept new solutions. We see industries that focus on working hours rather than efficiency on incentives and may be punished for work stoppages. If the market is large and capital-rich, then alternative strategies may be justified, such as Uber's global fight with regulatory and taxi unions.
All in all, in order to make full use of the huge opportunities brought by AI and ML, you should: 1. Avoid areas where large network companies have structural advantages; 2. Let products solve the obvious pain points of users, and the buyer does not have internal contradictions; According to the degree of willingness of the industry to adopt AI and ML technology, choose a target industry, and there is no major regulatory obstacle in this industry.
P03 Series Push Wire Connectors
Smaller size, space saving
Fast connection
111
Push-In Cable Connector,Dual Poles Quick Wire Connectors,fast connection cable connectors,cable managing connectors
Jiangmen Krealux Electrical Appliances Co.,Ltd. , https://www.krealux-online.com