Home>English>Marxist Research
John Case: AI — The Promise and The Threat
     Release time: 2024-04-11
  There is a lot to talk about on the subject of Artificial Intelligence (AI). But let’s cut to the conclusion first, in case the forest of diverse dimensions of this fairly deep topic gives you a headache!
  Liz Shuler, President of the AFL-CIO, hit the nail on the proverbial head when asked what to do about Artificial Intelligence. She said, for a start, “Strengthen collective bargaining rights, like, generally,” according to a recent Politico Interview.This may sound like an ‘automatic’ response from a union, but in fact it reflects and sums up serious and energetic work by the AFL-CIO leadership to engage the impact of AI on U.S. workers.Shuler goes further:“Through every industrial revolution, labor has been the force that has harnessed the technology and channeled it in a way that’s productive and safe. … We need to be that stabilizing force, and we are the only movement that can do that.” That’s a bold claim. But, I submit, it’s true.
  Let’s review some key events from the past year that demonstrate AI is a general purpose technology — meaning it will invade and transform every sector of the economy, whether service or manufacturing. And the economy is just a starting point for analyzing the cumulative effects: the place where the AI tech winners and losers can be measured in wealth, poverty, jobs, and economic security for families. These economic factors can only begin to address the broader political, cultural, war and peace dimensions of AI’s impact on workers’ lives and the society as a whole.
  The labor movement has always had a moral focus on “leaving no one behind” in industrial policy. Thus, the AFL-CIO’s participation by invitation in the presidential Select Committee on AI is focused on regulations and investments related to training workers for jobs in the event of displacement.
  In addition to a major expansion of collective bargaining rights, President Shuler also advocates for establishing a national “traffic cop” entity for evaluating new technologies. The entity would function similarly to the FDA (Food and Drug Administration) to make sure new products don’t kill people before they are released to the world. A people’s movement must be built to demand federal action and regulation around this initiative.
  Legal battles In bourgeois courts
  While the need for action and regulation is clear, the exact role of the “traffic cop” is not clear at all. Most of the “action,” so far, has been in the courts, where the issue is “intellectual property rights.” Daily, even hourly, millions of people post or share digital content on social media platforms that arguably violate some interpretation of what constitutes “property” in ideas, or other “intangible” products.
  It is the “intangibility” that causes a serious problem for “ideas” in the commercial, capitalist world. The famous economist Paul Samuelson once called ideas “crappy commodities” on two grounds: 1) it is very difficult to reduce the origin of any idea to a single person, especially in modern, highly socialized and collaborative institutions and environments. For example, up to 99% of patented pharmaceutical products consist of knowledge in the public domain. 2) Intangibles do not retain value like tangibles — a hammer, for example. And, they are copied, i.e. ‘stolen,’ and reproduced with relative ease.
  Despite these impediments, capitalism has little means other than “property” to protect or express commercial values. Thus, an immense legal edifice of patent and copyright laws and trade protection has evolved over the years. But the inherent contradictions of forcing ideas into fixed or rigid property molds is sending a wrecking ball into much of this edifice. There are many lawsuits working their way through the courts demanding that the major platforms pay compensation for violations of intellectual property rights or privacy rights.
  These lawsuits may or may not prevail. The AI engines or algorithms require truly immense data oceans to train and perfect AI models. If you think of the key role of oil in the industrial age, you have the key to thinking about data in the AI or cloud computing age. Only the current big platforms have the “oil” to deliver the anticipated performance. The lawsuits, and reform movements rising will need big (i.e. tax-the-rich-generated) resources to handle the enormous costs of AI transformations, and just transitions — way beyond the past, largely empty, talk. Getting money from the big tech companies will be a necessity. However, it is highly unlikely that they will be denied access to the data that people “willingly” shared on their platform.
  Further — without going into too many theoretical or technical details — the “crappy,” “leaky” attributes of intangible commodities, combined with the scale of data requirements, have contributed to a very rapid rise of monopoly firms in the AI industry. Operating within the basic laws of capitalist accumulation, the intangibility and scale of data requirements, and the AI tech companies profiting from them, act to counter and considerably weaken existing antitrust remedies to market concentration. If data “clouds” (i.e. groups of servers running software and databases that are accessed over the Internet) get smaller or fragmented, a competing company with a larger, less fragmented cloud gains a computational edge.
  The best remedy providing scale, economic justice, and international working-class unity (and the only real solution, in the long-term), is socialism. No more deadly adversaries — at least in near-earth spaces!
  The political side
  One might not buy that yet. So, let’s look at the political side of AI. In an important sense, AI is a reflection of our global, collective intelligence. Its powers of classification, simulation, prediction, noise reduction,computation, and other learning techniques appear to mimic our own learning processes. Only now, they are operating under the power of the owners of the cloud platforms.
  Facebook, for example, possesses billions of rich personal profiles. Included in such profiles are marketing preferences which serve the original advertising business models of Meta, Google, and Amazon, of course. But they also include detailed social and political profiles that are gold mines to anyone with a significant political interest, such as the billionaire owners of the “Magnificent Seven” AI corporations (i.e. Apple, Microsoft, Amazon, Nvidia, Meta, Tesla, and Google’s parent company Alphabet). Even with our great powers of imagination, we are likely to underestimate their ambitions and influential reach.
  Economic consequences are important, urgent, and immediate. However, the political dimension to the AI story is no less profound.
  President Biden, by executive order, recently granted U.S. national security agencies direct access to any “foreign or terrorist” sources or transactions accessing U.S.-based cloud computing resources — where Microsoft’s advanced AI computing capabilities are mostly concentrated, for example.
  One does not have to be paranoid to be alarmed about news of this intrusive state surveillance. It will thus be a key task of a socialist movement and government to establish democratic dominion over AI, indeed over development policies in general.
  The politics of AI can be summarized thus:
  It is disruptive. This has both an upside and a downside. On the one hand, it promises immense improvements in productivity. An AI-assisted healthcare worker may be able to effectively and accurately treat 10 times as many patients as before. Perhaps more.
  Disease detection and remediation are being revolutionized. Manufacturing will become even more automated. With that, the real costs of commodities should fall, as should the rate of profit (though not necessarily the absolute mass of profit) on productive capital. Many services and skills will be replaced with new tools. With increases in productivity, work time can be dramatically reduced and wages increased.
  AI-assisted education (including for the AI itself!) is already delivering high-level engineering, language, and other skills. One AI activity, centered around Stanford University, provides AI-assisted education to millions. I took the course myself for $50.
  These few examples are just the tip of the spear. In many industries, AI promises to enable the direct collaboration of any worker, anywhere in the world, with any other worker, anywhere in the world.
  The domination of media and politics by the billionaire, Wall Street, and military-industrial complex is vastly increased by AI. Like their domination of the S&P 500 stock index, the domination over the media by these “Magnificent Seven” leave very little room for influence byr the rest of us.
  Privacy and sovereignty
  Even under the most favorable outcomes, it is fair to ask: What privacy remains, either from corporate, state, or even not-so-hypothetical “independent” AI entities? The answer is, not much.
  I believe some of the inferences are clear. Whether or not privacy is being reduced may be difficult to answer. It currently varies greatly by one’s economic class. But it is definitely being redefined in generational practice, as are many if not most family and other forms of friendship, association, and community.
  At the same time, democratic global governance must advance, if the weaponization of AI is to be averted. The connectedness of global populations will increase. The interactions of global AI entities with the work of all peoples undermine many border lines.
  It is the work of the world that changes the world, every day.
  Which takes us back to the beginning with President Shuler and her call for expanding collective bargaining as the chief path available to working-class folks to defend themselves against disruptive technologies — historically, and now. If we do not separate these powerful technologies from billionaire control, it’s likely to be one billionaire dictatorship threat after another unto the horizon, until we do.
  The political reforms required to bring AI under democratic dominion seem a revolution away. But solidarity movements from the past year have been warming the blood of the labor movement. Asserting the principle of “leave no one behind” with every bit of force we can muster, is a match that, struck with good timing, can start a political prairie fire. Something AI is not yet trained to handle.
  Editor: Zhong Yao  Liu tingting
Related Articles