The 2024 Nobel Prizes in Physics and Chemistry had AI as a central focus. However, as much as this marks a landmark phase of recognition of this new set of technologies, the age old question of “what this tech will bring us”, comes up. As it turns out, we don’t have any accurate answer to this complex question.
More than a hundred years ago, a similar question was being asked in California, in the context of the development of the transcontinental railroad.
“What the Railroad Will Bring us,” was the title of an essay written by San Francisco Businessman Henry George, who wrote it in 1858[i]. This was an important question to be asking at this time, as are similar questions that we can ask, today. “What will AI Bring us,” in 2024.
Henry George was skeptical about the positive impact of the Railroad on the vast majority of the people. He wrote: “ The truth is, that the completion of the railroad and the consequent great increase and population, will not a benefit to all of us, but only to a portion. As a general rule (liable of course to exceptions) those who have, it will make wealthier, for those who have not, it will make it more difficult to get. Those who have lands, mines, established businesses, special abilities of certain kind, will become richer for it and find increased opportunities, those who have only their own labor will become poorer, and find it harder to get ahead – first because it will take more capital to buy land or to get into business; and second, because as competition reduced the wages of labor, this capital will be harder for them to obtain.” Kevin Starr, a historian of California seems to agree with this and history bears out that they both were right, at least in the short term.
As Starr adds “The next decade was one of recession, strikes and racial tensions in California. On the other side of the Atlantic, Marxism and revolutionary forces were creating other changes, that did impact the U.S., but not as significantly.” Railroad and infrastructure is a form of technology that can have a positive impact but also a deeply negative one, depending on whose land it is built on, who is displaced, what is the long-term impact of which towns/ cities are connected and which ones are left out of the development processes. This is a far more complicated question than simply saying “all infrastructure is good.”
Does AI really democratize data? A cursory glance challenges this assumption. Yes, AI does make life easy, in terms of data processing, carrying out some routine tasks and perhaps as the capacity of Generative AI grows, even more complex tasks. However, one must remember that AI technologies are capital intensive and are proprietary. This means that they will be deployed only where they are profitable. The recent attention of the Nobel Prize may of course bring more money into the space and while it may spur more innovation, the question of who is able to get into this space and who benefits from this tech is an altogether different question.
The question of ethical use of data is as urgent as ever. And while businesses are celebrating this as the arrival of a new era, there is reason to be cautious and also skeptical of many of the claims, as widespread use of AI can have very big ethical implications.
When one looks at other examples of how technology was supposed to make life easier and digitization is being positioned as the magic wand to solve all social problems, we see, in countries like India, that this is far from the truth. Making digitization and onboarding everyone on certain platforms or universal recognition systems is actually counterproductive and may leave many millions of people out, as this article points out.
The biggest challenge and problem with AI or any digital technology is the invasion of privacy and lack of one’s ability to “opt-out.” Further, there is also the question of authorship, giving credit to people’s intellectual property. As the Nature article points out “Yet, because the tools are trained on huge data sets — that often are not made public — these digital helpers can also clash with ownership, plagiarism and privacy standards in unexpected ways that cannot be addressed under current legal frameworks.” This is a very real concern and one that is not easy to answer. We are truly in the Wild, Wild, West moment in the development of AI and those who are too enthusiastic and jumping aboard this ship are better off slowing down and re-thinking this problem.
And of course, there are the ethical issues with it, too. One obvious one is the use of AI in surveillance technologies and also in warfare.
As The Palestine Laboratory, an award-winning book shows, how AI technologies have been used in wars, including the assault on Gaza, with untold suffering caused on civilians. These amount to war-crimes, as the ICC has ruled against Israel’s indiscriminate use of force against civilians. How much of the blame should AI get, for this?
The human element of using judgment and any pretense of “wisdom” is lacking in AI based systems. While AI based models can give us the most optimal decision based on certain parameters, they are still not as advanced to help us decide a fundamental human question of “What is the right thing to do.” For example, as the 972 investigative piece quotes an IDF personnel saying “We were not interested in killing [Hamas] operatives only when they were in a military building or engaged in a military activity,” A., an intelligence officer, told +972 and Local Call. “On the contrary, the IDF bombed them in homes without hesitation, as a first option. It’s much easier to bomb a family’s home. The system is built to look for them in these situations.” We see, in this comment, the manner in which an AI system was programmed to cause maximum damage and harm, with no regard for human life.
AI, it seems, is racist. And is causing immense suffering, as we speak. This is not a hypothetical, “what if” situation, but something happening now, as others have pointed out.
In his book The Palestine Laboratory, Antony Loewenstein points out that Israel has exported its technologies, including AI based ones, which are used primarily for repressing people, around the world. When we are outsourcing our judgement to a machine, it will just spit out what it has been fed – GIGO – Garbage in, Garbage out. This has been Israel government’s modus-operandi across the world, point out various commentators.
Is AI racist? There is emerging evidence that the models used to feed AI systems are deeply problematic. Shir Hever, an expert on Israeli arms exports is quoted in this book as saying “They say that these Europeans are so naive. They think that they can have human rights. They think that they can have privacy but that’s nonsense. We know that the only way to fight terrorism is to judge people by how they look and the color of their skin (p.16).” This is the foundation of some of the most destructive software, Lavender which is being used in Gaza, today.
If we are to objectively judge any technology, it is only fair to assess its impact on everyone- including those who are at the receiving end of its rewards or risks. And judging from its current state, there are many deeply problematic aspects of AI, that need to be addressed and corrected, before being deployed, at a mass scale.
Not to do so would be to increase one thing that seems to be increasing around the world – the repression of dissenting voices. And what better tool to do that than a faceless, morality free system.
[i] See full text at https://www.teachingcalifornia.org/resource/what-the-railroad-will-bring-us/
