Competition in Industries: The Good, the Bad, and the Future for AI
[November 2023] In recent weeks I have read or watched videos about the dangers of AI and in particular about regulating the dangers of AI. So, I thought I would use this as the basis of a new article. The examples in the article come from myself and others and have been selected because I feel they explain the issues in a manner many can relate to.
Competition is a driving force of progress in various industries, with both positive and negative outcomes. Let's take a closer look at the impacts of competition in automotive, pharmaceuticals, technology, and the looming challenges in the AI industry.
The Positive Side
Automotive: Competition among car manufacturers has made advances in safety and emissions. This benefits car owners with safer vehicles with better crash test ratings and less pedestrian harm, and contributing to a cleaner environment and better electric power units.
Pharmaceuticals: In the pharmaceutical sector, companies compete to develop life-changing drugs, treatments and vaccines, leading to cures and preventions of once life ending illnesses. This has been seen with recent COVID vaccines. This competition can also lead to lower prices when generic alternatives enter the market.
Technology: Competition between tech giants like Microsoft, Google, Apple, and Meta has resulted in powerful and user-friendly software, making information and services accessible to people at a low cost.
When the right incentive is there, win-win situations are possible.
The Dark Side
Automotive: Not all competition leads to positive outcomes. The infamous "diesel gate" that involved VW, Mercedes, BMW, and others manipulating emissions tests and compromising emission and environmental standards for profit.
Pharmaceuticals: Competition among US pharmaceutical companies in marketing prescription painkillers contributed to overprescribing and widespread addiction, causing a significant public health crisis.
Tech/Social Media: Competition among social media platforms for engagement and ad revenue has had negative consequences. Photo filters distort perceptions of beauty, leading to issues like body dysmorphia. Additionally, the competition for user attention has been criticized for privacy breaches and mental health concerns. I don’t really know enough about some of these issues but is worth remembering such tragic stories such as Millie Russell.
Lose-lose situations happen when individuals or companies prioritise self-interest, leading to harm for everyone.
A Cautionary Tale for AI
We can sometimes get so focused on the short-term goals that we lose sight of the big picture. We have a history of this, sometime knowingly and other unknowingly. Pollution and greenhouse gases, deforestation, and over use of antibiotics, are examples of how individuals and groups can get trapped competing in ways that ultimately hurt everyone – including themselves. The competition is often for short term profit.
The AI industry risks falling into this same trap. The race for computing power, funding, and talent may compromise safety in the development of such transformative technologies like AI and AGI (artificial general intelligence). Some industry leaders, like Anthropic, OpenAI, and DeepMind, have expressed the need to prioritize safety, but real action is needed.
To avoid catastrophe, AI leaders must move focus from short-term profits to the long-term benefits. Like safety standards in the automotive industry, we need to find measurable ways for companies to compete over safety criteria, alignment research, bias, etc., rather than pure technical capabilities. AI leaders should be willing and be encouraged to sacrifice individual advantages for collective benefit, ensuring responsible development and deployment of AI technologies. Good competition can deliver positive progress if incentives are aligned with the greater good – and benefit the AI technology companies in the process.
Doing Right by AI
If we get AI wrong, it would not be unrealistic to think this could lead to catastrophe. But, if done right, it could help solve many problems. I don’t know if this is a government, regulatory or industry thing – probably all of them. We have seen that with varied success we have been able to “regulate” away from danger. In the earlier examples, climate control agreements, regulation and social changes – sadly often driven by taxation (or tax benefits).
Consumers also hold the power to influence AI leadership. Choosing AI providers with a commitment to the greater good and asking critical questions can drive a positive change.
As technology evolves, it's essential to redefine the rules to harness the best in human nature for the future of AI and beyond. Consumers, industry leaders, and regulators play pivotal roles in shaping the responsible and beneficial development of AI technologies.
Related Articles:
