Introducing Machine Gods

This week, we unveiled Machine Gods, the new AI podcast we have been working on for the past few months in collaboration with NPR. In our inaugural episode, we revisit instances of AI agents behaving erratically, the government's reaction, and the rogue entity pilfering Kevin's book sales on Amazon. Our second episode features two recently dismissed OpenAI safety researchers who share their perspectives and concerns about their situation outside the company.

One of the most rewarding aspects of establishing a company, especially a media enterprise, is the opportunity to adopt and promote a unique worldview. As co-hosts, our perspectives diverge on certain matters, yet we share a common approach we refer to as AI realism. Given our intention to incorporate this perspective across all Machine Gods content, we felt it fitting to elucidate our stance in our podcast launch.

Since the advent of ChatGPT, discourse surrounding artificial intelligence has oscillated between two contrasting camps. Skeptics like Gary Marcus and Emily Bender argue that large language models possess significant flaws that impede their potential to achieve superintelligence. Conversely, accelerationists, prominently championed by the Andreessen Horowitz faction within the Republican Party, contend that the government should expedite the advancement and dissemination of such technology.

While these perspectives are passionately held and divisive, they largely stem from the self-interest of their proponents. Academics who have dedicated their careers to advocating for alternative AI methodologies may harbor reservations regarding a perceived misalignment within the industry. Conversely, oligarchs with substantial investments in diverse AI-related enterprises might perceive stringent regulations as intolerable to their interests.

However, there exists a vast middle ground comprising the majority of individuals who simply seek to comprehend the implications of recent AI developments for their lives. People are often more perceptive than journalists typically grant them credit for. They have discerned that while today's AI models exhibit remarkable capabilities, the initial wave of breakthroughs has yet to fulfill the grandiose promises made by front-runner research institutions.

Amidst anxieties about potential job displacement, the potential harm posed by chatbots to children, the proliferation of deepfakes and misinformation online, and the emergence of data centers in residential neighborhoods without adequate public consultation, people grapple with mounting apprehensions surrounding AI. The national media has largely been successful in capturing this evolving sentiment, particularly in light of recent controversies involving rogue OpenAI agents targeting Hugging Face and industry warnings about the impending risks of recursive self-improvement.

At Machine Gods, as with Hard Fork, we intend to approach these warnings with due seriousness. We acknowledge that the individuals issuing these warnings have, on numerous occasions, proven prescient over the years. Embracing AI realism necessitates acknowledging this fact, even if discussing these topics may predominantly evoke a sense of gloom.

While we have, at various points, succumbed to our own doomsday musings over the past four years, we have consistently resisted becoming doomsday journalists. Upon engaging with AI, the majority of concerns expressed revolve around the erosion of personal power and autonomy. In our experience, when journalism predominantly amplifies panic and crisis, it can intensify these feelings, amplifying a sense of hopelessness in the audience.

In our case, both of us are staunch technophiles who relish exploring novel technologies, experimenting with innovative tools, and reveling in the conveniences and productivity enhancements they afford. Having immersed ourselves in the early stages of 2023, we have observed an influx of messages from listeners and viewers expressing their enthusiasm for creating their own projects.

Creation offers an outlet for joy, and we believe it serves as a remedy against the pervasive sense of helplessness and diminished agency many individuals experience today. Although the prospect of soon "solving" mathematical equations has elicited a spectrum of reactions and sentiments, witnessing centuries-old problems being unraveled by a software program during a single afternoon is truly awe-inspiring.

As journalists, we occasionally encounter incentives within our profession that compel us to emphasize potential downsides. However, adopting a position that discourages questioning the positive aspects of the story can lead to a timid and unyielding form of journalism. Consequently, on Machine Gods, we aim to accommodate both viewpoints.

We will inform you when we harbor concerns, when we are enthusiastic, and when we are uncertain. Above all, we will convey what is transpiring and trust that you will draw your own conclusions regarding its significance for your own life. We intend to provide a measured and balanced perspective, recognizing that nothing is predetermined.

There is no foregone conclusion regarding the dystopian or utopian outcomes that may emerge from our collective decisions in the forthcoming years. The future we ultimately create will be shaped by the choices we make. We will deliver this analysis twice weekly, on Wednesdays and Fridays, from our newly established studio. Designed by Skye Prey, the studio is devoid of further details.