Generative AI in the cloud: Build or buy?

David Linsigao

Compiled | Yan Zheng

Produced by 51CTO technology stack (WeChat ID: blog51cto)

In technology There is an unwritten rule in the field: everyone likes to adopt other people's technology. But for many businesses, generative AI doesn’t seem to fit that mold. Generative AI is rapidly driving some critical decisions. Every organization faces an important choice: build a custom generative AI platform in-house or purchase a prepackaged solution (often offered as a cloud service) from an AI vendor.

DIY favors quantity and opportunity. It's weird, but the reason might surprise you. They might even lead you to rethink your enterprise genAI strategy An artificial intelligence platform can give enterprises complete control over its features and functions. AI technology can be adapted precisely to an organization's needs, ensuring compliance with a company's unique workflows and delivering a customized user experience. Note that DIY generative AI can be done on public, private, or traditional platforms. We are currently focused on open source implementations using specific genAI technology, either on-premises or in the public cloud

Natural language interaction does provide a more "human" way to handle static business processes. However, there are concerns that these systems could quickly become core to the business, and unless they have full control over all features and functionality, they run the risk of the systems not delivering overall value. In other words, if someone buys an AI platform with all the bells and whistles and it changes direction or even disappears, they will be at risk of a failing system and business failure

2. More money, more money More time, more risk

Building a complex generative AI platform requires a team of experts with specialized knowledge, but it is difficult to find enough experts in the existing talent pool. You'll need data scientists and AI engineers to work with cloud and non-cloud platform engineers to develop custom genAI solutions that meet the specific specifications of your business

Doing so may add complexity, and Expensive talent needs to be hired. I have a friend who is a CIO who sends his employees to graduation ceremonies at good technical universities and approaches them directly in the school parking lots to recruit people before they enter the open job market. It’s an approach that’s unsettling but also innovative

Most businesses need to be creative to attract enough talent. Due to talent shortages, some companies have to postpone projects or choose to purchase systems instead of building systems themselves

3. Purchase value

Purchasing this system can provide rapid deployment and development Out-of-the-box functionality. This means you can immediately use pre-built solutions for quick implementation. You can gain immediate value and accelerate time to market

Purchasing generative AI services comes with the importance of ongoing support, updates, and improvements. Although DIY methods can provide some help with some components, it is mainly necessary to build it yourself

Consider the cost of building and supporting the database before purchasing it from a database vendor cost comparison. Of course, AI systems are more complex and involve many more components, but the analogy is appropriate

The value of a build approach depends entirely on how you create a custom one-off based on your business needs Solution needs. Betting on additional cost, time and risk will pay off when you have complete control over the core system. For many, core systems will become an important part of the business, not just business automation

Using genAI correctly could become a critical factor in the success of a business in the coming years

4. Weigh all factors

When deciding to build or buy a generative AI platform, you need to consider all the pros and cons. First, building generative AI in-house can be costly. In contrast, off-the-shelf solutions offer practicality and cost-effectiveness

Building generative AI in-house requires assembling a skilled team, and off-the-shelf solutions give you access to the expertise of the AI ​​vendors who built the system. This means pushing the risk and cost to the vendor or provider

Finally, creating an AI solution from scratch means complete creativity and control over the technical process. This enables integration of compliance measures and exact functionality to meet requirements from the outset. We all know how construction works. Customization can lead to a lot of iteration and time-consuming development. Additionally, support and maintenance are critical for generating AI in-house. If this doesn't provide enough value to justify a do-it-yourself approach, consider purchasing, which eliminates the risk, time, and cost

We're going to see a lot of shit in the future decisions ultimately led to the bankruptcy of the company. Maybe it's because they bought the wrong equipment when they should have been building technology and weren't able to deliver the technology value the industry needed; or maybe because they lacked talent and limited budgets, they couldn't create valuable products

There is basically no doubt that the above situations will definitely happen

Please refer to the following link: https://www.infoworld.com/article/3711705/build- or-buy-cloud-based-generative-ai.html

The above is the detailed content of Generative AI in the cloud: Build or buy?. For more information, please follow other related articles on the PHP Chinese website!