The risks of building apps on ChatGPT

The topic of “AI” gets a lot of attention and press. Coverage ranges from apocalyptic warnings to Utopian predictions. The truth, as always, is likely somewhere in the middle. As developers, we are the ones that either imagine ways that AI can be use

Fly.io logo
Fly.io
•
Mark Ericksen
•13 min read•advanced•
--
•View Original
Serverless

Overview

The article discusses the risks associated with building applications on proprietary AI models like ChatGPT, emphasizing the importance of considering open source alternatives. It outlines various risks such as single provider dependency, regulatory changes, financial implications, and governance issues that can impact businesses relying on external AI services.

What You'll Learn

1

Why businesses should consider open source AI models over proprietary ones

2

How to evaluate the risks of relying on a single AI provider

3

When to diversify AI dependencies to mitigate risks

Key Questions Answered

What are the risks of building applications on proprietary AI models like ChatGPT?
The article identifies several risks including single provider dependency, regulatory changes that could affect service availability, financial risks associated with AI service costs, and governance issues stemming from leadership changes within AI companies. Each of these factors can significantly impact the reliability and viability of applications built on such services.
How can businesses protect themselves from AI service interruptions?
To mitigate risks from AI service interruptions, businesses should consider using open source AI models that can be self-hosted. This approach allows for greater control over the AI integration, reducing dependency on external providers and protecting against service outages and policy changes that could disrupt operations.
What financial risks are associated with using proprietary AI services?
Proprietary AI services often have unpredictable costs, which can jeopardize business models if the pricing structure changes unexpectedly. The article highlights that many AI chatbots, including those from leading providers, can lose money on each interaction, raising concerns about long-term sustainability and profitability.
What governance risks should businesses consider when using AI services?
Governance risks include potential instability within AI companies, as exemplified by the leadership turmoil at OpenAI. Such changes can affect the direction of the service and introduce uncertainty, making it risky to rely heavily on their offerings for critical business functions.

Key Actionable Insights

1
Consider transitioning to open source AI models to reduce dependency on external providers.
By self-hosting AI models, businesses can avoid service interruptions and maintain control over their applications, ensuring that critical features remain operational even if external services face issues.
2
Evaluate the criticality of AI dependencies in your applications.
Understanding how essential AI components are to your business can help in making informed decisions about whether to rely on proprietary services or seek alternatives that offer more stability and control.
3
Stay informed about regulatory changes that could impact AI services.
As the AI landscape evolves, keeping abreast of potential regulations can help businesses anticipate changes that might affect their operations and adapt accordingly.

Common Pitfalls

1
Relying too heavily on a single AI provider can lead to significant operational risks.
If the provider experiences downtime or changes their service model, businesses may find themselves unable to deliver critical functionalities, leading to potential loss of customers and revenue.
2
Underestimating the impact of regulatory changes on AI services.
Businesses may not consider how government regulations can suddenly alter the landscape of AI offerings, potentially rendering their applications ineffective or non-compliant.

Related Concepts

Open Source AI Models
AI Service Dependencies
Regulatory Impacts On Technology