A handful of IT companies hold most of the tech world in their hands: search engines, cloud platforms, basic software for computers and other electronics are made by companies that can be counted on the fingers of both hands. The same is true of the development of artificial intelligence. It's simple – they have the resources, they have the data.

At the end of November, when the epic saga of the collapse of OpenAI's board of directors unfolded, the casual observer might well have believed that the generative AI industry was a vibrant competitive ecosystem.

But this is not the case – nor has it ever been. And understanding the causes is fundamental to understanding what artificial reason is and what threats it brings. Put simply, in the context of the current paradigm of building larger and larger AI systems, there is no AI without Big Tech technews.bg.

Chaos in OpenAI: what's going on in the company that created ChatGPT?

Big Tech has an advantage

With painfully few exceptions, every start-up, new entrant, and even an artificial intelligence research lab depends on these firms. All rely on the computing infrastructure of Microsoft, Amazon and Google to train their systems. They also rely on the huge scope of the consumer market of these same companies to implement and sell their AI products.

Big tech firms have gained significant advantages over the past decade. Thanks to their dominance and business models, they now own and control the ingredients needed to develop and deploy AI on a large scale. Large IT firms are also shaping the incentive structures in AI research and development, defining the present and future of the technology.

The recent OpenAI saga is revealing. In it, Microsoft exercised its quiet but firm dominance over the entity at a "limited profit." It is clear that he who pays, he orders the music.

The Risks of Concentration of Power

Concentrated power is not just a problem for markets. Relying on several corporate structures to build basic infrastructure – this is a problem for democracy, culture, society and the individual. Without significant intervention, the AI market will only reward and establish the same companies that have reaped solid profits from their invasive business models, commercializing everything on the Internet, often at the expense of the public.

End of Drama: Altman to return to OpenAI as CEO, but with new board of directors

The Cambridge Analytica scandal was just one of many cases that exposed this ugly reality. Concentration also creates single points of potential failure, which poses real security threats. U.S. Securities and Exchange Commission Chairman Gary Gensler warned recently that having a small number of AI models and actors at the heart of the AI ecosystem poses "systemic risks to the financial order."

But the attempted expulsion — and subsequent reintegration — of OpenAI co-founders Sam Altman and Greg Brockman didn't just demonstrate Microsoft's power and influence and bring the issue into the spotlight. It also proves that these trade agreements give Big Tech absolute control over the trajectory of artificial intelligence.

Microsoft already has a seat on the OpenAI board, albeit without the right to vote. But the real leverage that Big Tech holds for controlling the landscape in AI is the combination of computing power, data, and huge market reach. To follow its current approach to developing AI, OpenAI had to strike a deal. The company exclusively licenses its GPT-4 system and all other OpenAI models to Microsoft in exchange for access to Microsoft's computing infrastructure.

For companies hoping to create basic models, there is no alternative to working with Microsoft, Google or Amazon. And everyone busy developing AI is aware of this.

Programmer: When AI surpasses us in intelligence, it has to like people, otherwise it will destroy us

It is also indicative of what came out of Altman's decision to seek investment from Saudi Arabia and the Emirates in a hardware plant in the hope that it would rival Nvidia. This company has almost a monopoly on state-of-the-art AI training chips. This is another key point in the AI supply chain. U.S. regulators have driven away Saudi Arabia's initial investment in the Altman-backed company called RainAI.

Alternatives

There are several meaningful alternatives, even for those who wish to go the extra mile to build industry-independent AI. As already rumored, "open source AI" — an undefined term that is currently used to describe all sorts of models and systems — cannot, on its own, offer a quick break from the grip of Big Tech.

On the one hand, many open-source AI projects work through computing credits, revenue-sharing, or other contractual arrangements with tech giants struggling with the same structural dependencies. On the other hand, Big Tech has experience in ways to squeeze profit even from open source developments. Open source AI can offer transparency, reusability, and scalability — and that's fine. But it does not solve the problem of concentrated power in the market of artificial reason.

The OpenAI-Microsoft saga also demonstrates another fact that often remains unseen in the AI hype: there is still no clear business model beyond increasing cloud profits for Big Tech by merging AI services with cloud infrastructure. Microsoft is not the only example. The situation is similar with Amazon, which recently struck a $4 billion investment deal with Anthropic. There are other examples.

Visibly, Big Tech is becoming more assertive in implementing deft maneuvers to protect its power over the market.

Will regulations help?

Regulations? Yeah, that might help. But government policy often leads to the affirmation, rather than mitigation, of the power of these companies. And how else can it be, since they use their resources and their access to political influence. We can easily track how each of these large IT companies has made "investments" amounting to millions and even billions of dollars in most countries where they are present. Politicians take it well, they praise it. They talk about these investments as a great benefit.

EU agrees rules governing the use of artificial intelligence

We can also see how much money Big Tech pours into lobbying directly – at least for the EU there are such statistics. It's not much different elsewhere. The technique gives companies reason for artificial intelligence to ask for tolerances. Now the EU has passed its own law on artificial reason, and tech industry giants are already bypassing their governorates, as similar regulations will soon appear in the US and elsewhere.

We need to introduce uncompromising transparency obligations that leave nothing hidden around fundamental issues like the data that AI companies have access to to train their models. We also need liability regimes that burden companies to demonstrate that they meet basic privacy, security, and distortion avoidance standards before their AI products are released to the public.

And to combat over-concentration, we need compelling regulation that requires business separation between different layers of the AI ecosystem — and prevents Big Tech from leveraging its infrastructure dominance to cement its position in the AI model and app market.

Follow the channel on

artificial intelligence

Amazon

Google

data

software

Artificial Intelligence Regulations

Search engines

Cloud platforms