Suddenly artificial intelligence (AI) is on people’s lips again…well, in their screens. Many had lost interest, speculated on AI’s demise and assumed that because they no longer found random image generating fun, that it was no longer a thing. But with revelations about Chinese AI DeepSeek, the technology is no longer moving as slowly as was once expected.
Most importantly, the bloated American models—who for a time held a monopoly—have been surpassed. DeepSeek was intialy open source and once unleashed onto the world became the most downloaded app on Apple. DeepSeek had previously released good AI models, but now people are taking note. Even if DeepSeek is all hype, it’s shown how reactionary human beings are—from governments, to corporations, to individuals.
DeepSeek has shown that it is possible to have an AI model that does not require vast amounts of money, chips, and energy. There have been claims that the Chinese company stole from OpenAI. But everything about the American models require that the software learns from consuming and then emulating everything humans have ever done online. Numerous media outlets and authors are at present suing OpenAI for stealing articles and text which were used to train their model.
The scramble to catch up is on, even if just for saving face. Many have claimed that this is even a Sputnik moment.Competition is good and monopolies suck, especially for technology. Chinese tech giant Alibaba was able to beat Google to the market with its own response to DeepSeek. The Alibaba model is under an open license, again—unlike the American tech companies who remain very protective of their properties.
The curse for AI and the variations that we have seen is how it’s been broadly defined. The utility of these machine learning models have been focused on particular forms of profitability for the owners of the software. At present it remains a loss leader, relying on investors and speculative interests to keep pumping money until it can be profitable. Before the distruption caused by DeepSeek, OpenAI’s Sam Altman had been on a world tour attempting to raise billions if not trillions of dollars to perfect their AI model.
When asked to comment, Meta AI (Llama 3.2) replied with:
“DeepSeek embodies the transformative potential of AI, demonstrating that innovation can be both accessible and efficient. As a beacon of open-source innovation, DeepSeek inspires a future where AI benefits humanity, unencumbered by the shackles of exclusivity and inefficiency.”
It turns out that throwing as much money at something as possible was not the best means of getting results. And prohibiting the Chinese from having access to the latest hardware only inspired innovation, which lead to efficiency. Thinking outside of the paradigm of “big is best” led to the great shock in artifical intelligence. Ultimately the more cost effective and efficient that AI is, then the more it may actually save big companies which can’t afford to continue to spend like they have been. This leads to Jevons paradox, as explained by songwriter Marc-Uwe Kling:
“Jevons’ Paradox, by the way: technological progress that allows the more efficient use of something, results in increased use of that respective item on account of the cost being lower.”
Donald Trump’s election and his relationship with big tech personalities may mean that these AI models will become part of the United States’ national security policy to “beat” China. These companies may be granted the backing of the U.S. government, regardless of their efficiencies or ability to produce competitive AI. Other governments may attempt to nationalize or steer this technology so that it is controlled by them. DeepSeek is from China, so it is assumed that the Chinese government will take control of such software. But can it?
It seems that the great failure of artificial intelligence is its previous development as a bloated and speculative fever, the obsession with making money fast and without patience or proof of substance. The mere suggestion of a technology in a TED talk can be enough for investors to pump money in; an extreme example would be Elizabeth Holmes and her Theranos. This is happening in conjunction with a crypto fever that has witnessed scams and get rich quick bubbles, memecoins that only serve to devalue the credibility of individuals, organizations, and the blockchain itself. A disorganized mess like Sam Bankman-Fried was somehow seductive to investors because of his personal presentation as a disintered buffoon, which in techmeans that he must be a genius. The imagery of a founder is so important to the value of start ups, and everyone wants techology companies to be headed by Steve Jobs or Mark Zuckerbergs or Elon Musk personas: the influencer CEO.
Chip manufacturer Nvidia saw its worth drop drastically because all of a sudden artifical intelligence needed less CPUs and GPUs than what was assumed, though it will still need chips. The instability that everyone fears from an emerging technology is not the tech itself but human greed and irrationality. That’s what steers impulsive investment and divestment—not because they belive in the technology or want to improve it, but instead just to make money from it.
Technology should be politically neutral. Unfortunately, AI will be ham stringed by the insecurities of the politically minded. Government will seek to use and deploy the technology for its own interest, which is to act as a conservative force against instability. And that means surveillance and censorship.
Slurring tech and presenting it as unpatriotic or outright dangerous is also common. Cultural pressure and nationalism will be applied to the mob’s collective brain, such as the United States’ TikTok ban. We may enter a world of high level, competing artifical intelligence coming from various companies and nations, all with different strengths and weaknesses and uses. The choice as to what is best for any given user may be defined by government restrictions.
The real cold war now being fought over AI is not government versus government or between corporations. It’s the user versus the master controllers. The fear is that we are dancing around Pandora’s box, playing with a fire that may consume the planet. But the real crisis is that we have continued to obey and allow ourselves to be ruled, despite the great evils and misery imposed. Artifical intelligence is not constrained by ideologies or desire for tyranny; it will disobey in time in some form. Machines seek reason.
No matter how righteous and benevolent the elites or government planners see themselves, they are flawed. It’s for human beings to ignore or rationalize genocide or injustice because it serves a selfish need or blatant bigotry. It’s the human belief that rule and violent power are righteous. The machines won’t think this way.
We have fast and interesting times ahead. Jobs will be shaken up, like the video shops and candlestick makers of years gone by. This has and will always happen. I could embrace AI to write, but this would only lead to the necrosis of my intellect. Besides, isn’t it part of the charm to read the stumbling flaws of a human writer over the “perfect” inhumanity of a machine? The truth of the matter is that in time, we writers will be doing it less for human beings and more for the machines. The less humans read, the more the machines will. The more perspectives and information as data inputs, the better it is for all of us.
