If you think all critical voices in the AI debate today are tech pessimists, think again. Scientist and best-selling author Gary Marcus proves that the most genuine tech optimist is today also the most critical of the current state of affairs. This AI scientist says that AI is not living up to its full potential, but is instead thrust ahead by a few greedy powers that are twisting and turning to rush through a deeply flawed ghost version of what AI could have been. Generative AI is a race between a couple of big tech companies and their leaders without consciousness. The FOMO they are creating narrows our attention down to the most basic, but also most dangerous, form of AI invention
The book is divided into three parts: In part 1 Marcus dissects the current state of affairs in AI business and politics. He traces the core social and ethical issues with AI that we are facing today to the hype created around Generative AI – an evolution he describes as deeply flawed and the history of which is characterised by the interests of a few big tech companies. This depiction leads to the second part of the book’s to-the-point and deeply honest depiction of the power and money behind the powerful public rhetoric leaving the world in a sense of urgency to adopt and implement generative AI with little or no social impact assessment. In the third part of the book, Gary Marcus takes us through recommendations for action for a more responsible AI future: Data Rights; Privacy; Transparency; Liability; AI Literacy; Independent Oversight; Oversight Comes in Layers; Incentivizing Good AI; Agile Governance and the Need for an AI Agency; International AI Governance; Research into Genuinely Trustworthy AI. Naturally, a major focus is in this part of the book on US internal politics. Most likely the mitigation of the consequences of Silicon Valley Big Tech may also mostly happen through internal US market strategies, law and politics. Marcus also recognises the pivotal role of international AI governance (through intergovernmental organisations like the UN), and importantly the EU in paving the way with regulation and responsible innovation (while also warning us about the immense lobby powers that are always lurking where we least expect it). Calling for the establishment of a legally binding international AI Treaty in the book, this year’s adoption and opening of the Council of Europe’s Convention on Artificial Intelligence is an important reminder to hold governments accountable (see more about the convention here https://www.caidp.org/resources/coe-ai-treaty/). The convention has by now been signed by e.g. the EU and the US and others might want to follow Norway’s example, sign it, and make it count for also the private sector.
Gary Marcus’ Taming Silicon Valley – How We Can Ensure That AI Works for Us” reads as a manifesto for action to take back world control from Silicon Valley.
Read it and weep! … and then act. As Gary Marcus puts it:
“We can push for a better, more reliable, safer AI that is good for humanity, rather than the premature, rushed-out-the-door technology we have now that is making a few people very rich, but threatening the livelihoods of many others.”
