OpenAI: A Blessing or a Curse
Are OpenAI and artificial general intelligence as great as people think? We break it down to discuss the challenges on the road ahead for AI implementation.
Join the DZone community and get the full member experience.
Join For FreeThe human desire to avoid mundane chores and simplify all tasks is as strong and persisting as humanity itself.
Through the ages, this desire manifested itself in many forms. Some, such as slavery or serfdom, were definitely cruel and only served the very few. Later on, the aristocracy went to great lengths to make their lives the best they could be, relying on an entourage of servants to tend to their every need. As the Industrial Revolution introduced the miracle of large-scale manufacturing and eventual automation, we finally discovered a way forward that would not rely on human workers.
Little did we know that the twenty-first century would bring an entirely new tool as potent as the stuff of fairytales to transform our lives forever.
- Humankind always strived to automate mundane tasks.
- Artificial general intelligence is the revolution in automation that dictates how we live and work.
- GPT-3 language model is the latest incarnation of AI that can process and simulate human language with unprecedented accuracy.
Less than four years ago, two prominent visionaries came together to give us the fruit of the future — OpenAI and its artificial general intelligence. Founded by Elon Musk, Sam Altman, and others, this fountain claimed to research and develop artificial intelligence, most likely for the benefit of humankind. OpenAI is a non-profit organization that makes API available to all developers who want to utilize artificial general intelligence in their software. Now, let’s focus on the promised benefits of AI.
The most promising and intriguing result of the OpenAI foundation that was rather promptly delivered is the GPT-3 AI autoregressive language model that steers the scientific and the general public over its potential implications. GPT-3 can simulate human speech and writing in an astonishingly sophisticated manner. Recent demonstrations of its capabilities show that AI can answer all sorts of questions with impressive accuracy and ease. Even humor and creativity are seemingly within its grasp.
The AI can write coherent and believable copy based on just a few initial sentences. GPT-3 can summarize complex documents, approximate the sentiment of a text, and even generate ideas or guess your age.
Surely, these capabilities may sound alluring to a range of professionals whose line of work revolves around writing copies, generating slogans, and so forth. Not to say that GPT-3 cannot come in handy to almost any professional at least in some capacity. Why not delegate your basic tasks to the AI and save time, right?
And yet, this apparently revolutionary AI doesn’t come without causes for concern and controversy. What’s interesting, after the initial trials, was that the AI was deemed too dangerous for public release by its own creators. The Verge article also stressed the “promising and problematic” nature of GPT-3.
Let’s explore these concerns to see what are some of the unforeseen implications this AI may have upon its public release.
- There are serious ethical issues related to the integration of AI.
- AI-based language models can pose as real humans, misleading their interlocutors.
- AI can also spread misinformation and biases since it may not be taught to discriminate against them.
- Automation using AI can leave millions of people jobless and as such should be regulated and taxed.
Ethical Issues and AI Posing as Humans
Some of the AI use cases can raise serious ethical questions. For instance, communicating with an advanced text or voice chatbot can feel surprisingly real and almost indistinguishable from talking to someone. Unless the conversation goes terribly wrong, that is. But is this level of believability so good when it comes to AIs posing as fictional TikTok influencers or chat pals from various apps popular with Gen Z audiences?
Such real-life applications of GPT-3 and similar technologies remind me of many films that depicted the appeal and danger of AI. For an old-school millennial like me, the line between creative potential and outright deception is blurry at best. Another movie comes to mind where machines rose against their human creators, but that is hopefully a different story.
Misinformation
There is also a growing fear that AI can spread misinformation and biases. Much like LeeLoo from The Fifth Element, who learned everything about human society from online databases, advanced AIs browse the internet to learn and build their knowledge. Depending on what information sources AIs build their neural networks upon, their understanding of our cultures, social norms, and communicative patterns may vary greatly.
GPT-3 is fed countless forum threads from Reddit, Quora, and other popular platforms. Consequently, the AI may not discriminate between what it comes across and treat falsehoods much like facts. Thus, all human biases, conspiracy theories, and lies can proliferate if left unchecked.
Impact on Jobs, Businesses, and Governments
Automation is yet another field to see a significant boost relying on the power of AI. The number of industries and professions where automation is becoming a popular and viable solution is constantly growing. However, more people are laid off as a result of this process. Potentially millions of households can lose income and experience significant financial hardship if automation proceeds at this pace. Besides, governments will receive fewer taxes from businesses, which will affect the stability of national and global economies.
Speaking of automation and taxes, Bill Gates states that robots should be taxed in his interview with Quartz. Businesses that install robots and practice automated production while laying off human labor should be subject to higher taxation. The income from these taxes should help finance jobs where human workers are essential. For example, caretaking of the aging population is seriously underfinanced.
Furthermore, Gates suggests that governments should step up their involvement in this process to offset deepening inequality created by automation. Otherwise, people, especially those whose job positions are subject to a robot replacement, are likely to feel resentment rather than enthusiasm for AI innovation.
To Sum It Up
OpenAI and its GPT-3 language model represent a significant technological transformation poised to have profound effects on our societies. AI can help make our lives better, safer and our industries more productive. The freed time and resources can help us concentrate on tasks where only human creativity and empathy matter. However, there are numerous ethical pitfalls, concerns of job loss, and growing inequality that can become even more serious if AI is not regulated. It is time for us to discuss and decide what we should do to make AI truly a blessing, not a new curse.
Opinions expressed by DZone contributors are their own.
Comments