The future of generative AI and its ethical implications

Generative AI is revolutionizing the way we experience the internet and the world around us. Global AI investment rose from $12.75 million in 2015 to $93.5 billion in 2021, and the market is expected to $422.37 billion by 2028.

While this view may make it sound like generative AI is the “silver bullet” for moving our global society forward, it comes with an important footnote: the ethical implications are not yet well defined. This is a serious problem that can hinder further growth and expansion.

What generative AI does right

Most generative AI use cases offer cheaper and more valuable solutions. Generative Adversarial Networks (GANs), for example, are particularly well-suited for promote medical research and accelerate discovery of new drugs.

It is also becoming clear that generative AI is the future of text, image and code generation. Tools like GPT-3 and DALLE-2 are already widely used in AI text and graphics generation. They have become so good at these tasks that it is almost impossible to distinguish human-made content from AI-generated content.

The million-dollar question: What are the ethical implications of this technology?

Generative AI technology is evolving so fast that it already exceeds our ability to imagine future risks. We must address critical ethical questions on a global scale if we are to stay ahead of the competition and see long-term sustainable market growth.

First it is important to briefly discuss how foundation models such as GPT-3, DALLE-2 and related tools work. They are deep learning tools that essentially try to “outperform” other models by creating more realistic graphics, text, and speech. Then labs like OpenAI and Midjourney train their AI on huge data sets of billions of users to create better, more sophisticated outputs.

There are plenty of exciting, positive uses for these tools. But we would be remiss as a society not to recognize the possibility of exploitation and the legal gray areas this technology exposes.

For example, there are currently two main questions under discussion:

Should a program be able to attribute the results to itself, even though the output is derived from many inputs?

Although there is no universal standard for this, the situation has already arisen in the legal field. The US Patent and Trademark Office and the European Patent Office have rejected patent applications filed by the “DABUS” AI developers (who are behind the Artificial Inventor Project) because the applications cited the AI ​​as an inventor. Both patent offices ruled that non-human inventors are ineligible for legal recognition. However, South Africa and Australia have ruled that AI can be recognized as an inventor on patent applications. In addition, New York-based artist Kris Kashtanova recently received the first US copyright for creating a graphic novel using AI-generated artwork.

One side of the debate says that generative AI is essentially a tool that can be wielded by a human creator (such as using Photoshop to create or modify an image). The other side says the rights should belong to the AI ​​and possibly its developers. Understandably, developers who create the most successful AI models want content creation rights. But it is very unlikely that this will work out in the long run.

It is also important to note that these AI models are reactive. That means the models can only “respond” or produce output based on what they are given. Again, that puts control in the hands of people. Even the models that are left to fine-tune themselves are still ultimately driven by the data people give them; therefore the AI ​​can’t really be an original creator.

How do we deal with the ethics of deepfakes, intellectual property and AI-generated works that mimic specific human creators?

Humans can easily become targets of AI-generated fake videos, explicit content, and propaganda. This raises questions about privacy and consent. There is also a looming possibility that people will be out of work once AI can create content in their style with or without their consent.

A final problem arises from the many instances where generative AI models are consistently rendered bias based on the datasets they have been trained on. This can further complicate the ethical issues, as we need to remember that the data used as training input is the intellectual property of someone else, someone who may or may not consent to their data being used for that purpose.

No adequate laws have yet been written to address these issues surrounding AI output. However, if it is generally determined that AI is just a tool, it follows that the systems cannot be responsible for the work they create. After all, if Photoshop is used to create a fake pornographic image of someone without permission, we blame the creator and not the tool.

If we assume that AI is a tool, which seems the most logical, then we cannot attribute ethics directly to the model. Instead, we need to look deeper into the claims made about the tool and the people who use it. This is where the real ethical debate lies.

For example, if AI can generate a credible thesis for a student based on a few inputs, is it ethical for the student to pass it off as their own original work? If someone uses a person’s image in a database to create a video (evil or benign), does the person whose image is used have any say in what is done with that creation?

These questions only scratch the surface of the possible ethical implications that we as a society need to work out to further develop and refine generative AI.

Despite the moral debates, generative AI has a bright, boundless future

Right now, IT infrastructure reuse is a growing trend fueling the generative AI market. This lowers barriers to entry and promotes faster, more widespread adoption of technology. Because of this trend, we can expect more indie developers to come up with exciting new programs and platforms, especially when tools like GitHub Copilot and Builder.ai become available.

The field of machine learning is no longer exclusive. That means more industries than ever can gain a competitive advantage by using AI to create better, more optimized workflows, analytics processes, and support programs for customers or employees.

In addition to these advancements, Gartner predicts that by 2025 minimum 30% of all new drugs and materials discovered will come from generative AI models.

Finally, there is no doubt that content such as stock photos, text, and program coding will shift to being largely AI-generated. In the same vein, misleading content will be more difficult to distinguish, so we can expect the development of new AI models to counteract the proliferation of unethical or misleading content.

Generative AI is still in its infancy. There will be growing pains as the global community decides how to deal with the ethical implications of the technology’s capabilities. With so much positive potential, there is no doubt that it will revolutionize the way we use the internet.

Andrew Gershfeld is a partner of Flint Capital.

Grigory Sapunov is CTO of Inten.to.