By JEFFERSON CHOMA
In the last weeks, there has been a lot of talk about the dangers of Artificial Intelligence (AI), especially because of the impact of the launch of ChatGPT, an artificial intelligence chatbot developed by the company OpenAI and launched in November last year.
The application belongs to the new generation of AI, capable of creating text from machine learning systems and large language models (LLM). In short, the application “learns” from searches in open databases and also by analyzing user-generated stimuli. In the case of ChatGPT, the application responds to the user as if they were a real person, developing a dialogue on any topic, including itself.
A few days ago, OpenAI announced ChatGPT-4, a new version of its chatbot that can now generate more complete answers, more closely resembling text written by humans, and can also analyze images. But this version is paid.
There are also other image platforms, such as Dall-e, Midjourney, and Stable Difusion, which produce realistic scenes from user-submitted definitions. On the internet, images of Pope Francis wearing a puffer jacket went viral. Some said he had hired a stylist, but the image was nothing more than a rendering made by the Midjourney AI. Selfies taken by historical personalities such as Jesus Christ, Napoleon Bonaparte, and Cleopatra were also published. All of these were made by Artificial Intelligence.
Aside from the awe surrounding the new technology, there are a number of concerns. Italy’s Data Protection Agency has provisionally blocked the use of the ChatGPT platform across the country, citing the “absence of any legal basis justifying the massive storage of personal data to ‘train’ the algorithms that guarantee the platform’s operation.”
Another controversial episode was the launch of an open letter calling for a moratorium on AI research. The document, signed by a group of entrepreneurs, academics, and researchers, among others, warns that the development of AIs “may bring profound risks for society and humanity” and that the technology “should be planned and managed with the corresponding care and resources.”
One of the signatories of the letter is Elon Musk, owner of Tesla and the richest man in the world. He has collected an ample portfolio of controversies himself, such as defending a coup in Bolivia to gain access to lithium, an important metal for the manufacture of batteries for Tesla’s electric cars. He also downplayed Covid-19 at the height of the pandemic.
It is naive to believe that Musk is really concerned about “ethics and accountability” in the development of AI research. Actually, there are other reasons for doing this. In 2018, Musk failed to take control of the company. And today, most of OpenAI’s research is funded by Microsoft, a company that competes with Musk in producing electric vehicles and developing innovative technologies. Microsoft provided OpenAI with a $10 billion investment for exclusive access to GPT-4, which would power its own Prometheus model for its search site Bing.
Does AI pose a threat to humanity?
But, with all this, what are the possible impacts of AI on society? Does it pose a threat to humanity? As in Hollywood movies, can machines rebel against humanity and take over?
It is not possible, within the limits of this article, to analyze this new technological tool’s impact in all its dimensions. However, aside from science fiction, this is not the greatest danger of the development of AI research. After all, as Margaret Mitchell, one of the most important AI researchers, reminds us, “machines are like parrots who repeat the nonsense taught to them by their owners.”
So the problem is what and who teaches the machines to repeat the nonsense. Today all AI research is in the hands of a few private companies, such as Microsoft, Google, IBM, Amazon, Apple, Facebook, China’s Baidu and Alibaba, and India’s Infosys. The scope of application of AI products is infinite. It ranges from gaming technologies to e-commerce, data analytics, and facial and voice recognition. And, of course, chatbots can help spread disinformation and even fake news, as well as facilitate the spread of cyberattacks. ChatGPT itself makes numerous mistakes when it “invents” information.
How can we guarantee that in the future chatbots will not manipulate the accounts of social facts and history itself, as described by George Orwell in his dystopian novel “1984”? Indeed, this is already happening with search sites under dictatorial regimes. In China, Google searches are censored and access to sites considered politically sensitive or containing information contrary to the interests of the state are blocked. In 2018, it was revealed that Google itself is developing a version of its search engine adapted to China’s censorship. This is just one small piece of evidence that the company is willing to do anything to make a profit.
Even in so-called “democratic” countries, responses for certain types of searches depend on who spends the most money on sponsorship and advertising, which influences the ranking of search results on the internet. This means that companies with more financial power may have more visibility in certain search results.
How can this sector be regulated? Who will enforce the laws in the face of the powerful interests of these corporations? The fact is that, despite timid regulatory measures, the influence of money on the internet is already a reality.
Technological progress and capitalism
We are facing a new wave of innovations that is replacing the workforce with machines at an unprecedented pace. This is taking the form of increasing automation in the most diverse activities. In addition to Artificial Intelligence, the development of Industry 4.0, the Internet of Things, Cloud Computing, 3D Printing, and the development of autonomous vehicles will be an inevitable reality in the coming years, which will likely be applied to everything from the assembly of a simple sandwich to the manufacture of the most delicate sheet of paper.
It is estimated that more than 50 million jobs in the United States are directly threatened by increasing automation. In the rest of the world, the estimate is that between 400 million and 750 million jobs will be replaced in the next 10 years.
The development of science and technology is a product of social relations. This means that it has a clear purpose: to serve capitalist accumulation. Capitalism demands permanently revolutionizing the social and technical conditions of the labor process in order to guarantee a greater appropriation of surplus value. But under this system, technology serves to create unemployment, reducing jobs and production costs of the capitalists through the widespread use of applications that result in precarity.
The most peculiar thing, without a doubt, is that the current technological development would already allow a drastic reduction of the working day, an almost total liberation from hard labor, and the inclusion in the production process of all unemployed people. However, this requires a radical transformation of the whole of society and its productive structure, building a socialist society that completely reverses its logic.
In a socialist society, technological and scientific development would become an extremely important tool to be applied in the reduction of the working day and in the full dissemination of information. It would increasingly allow the full participation of the population in political and economic decisions, facilitating the planning and monitoring of the implementation of the goals set. Imagine, for example, the use of the internet as a tool for the flow of information about production. Work would take place in a truly collaborative and communitarian environment, which now only occurs in small spaces and for the benefit of private companies. It is not difficult to imagine the benefits of Industry 4.0 or AI systems applied to the welfare of humanity.
In Marx’s words, it would put an end to “the life-long repetition of one and the same trivial operation, […] that thus reduces him to the mere fragment of a man.” Instead, it would free humans from the brutality of alienated labor and cultural backwardness, since “the universality of their development, their amusement, and their activity depends on the economy of time. The economy of time: this is what every economy finally boils down to.”
Originally published on www.pstu.org.br, 4/12/2023.
Translation: John Prieto