The 3-Year AI Deadline: How to Stay Relevant (2026-2028)

David and Daniel Lieberman, founders of Gonka, emphasize that the future reality hinges on who controls access to artificial intelligence. They highlight a scenario where big tech companies gatekeep AI access, contrasting it with an ideal where AI is freely accessible and affordable for everyone, leading to increased productivity.

The conversation raises concerns about how AI-powered tools like ChatGPT are changing information consumption habits. Instead of conducting independent research across multiple sources, people increasingly rely on a single source (like ChatGPT), potentially leading to consequences and a decline in critical thinking skills.

The Liebermans illustrate AI’s potential with an anecdote about using ChatGPT to diagnose a plumbing issue. Instead of costly replacements, the AI identified a missing part, demonstrating how AI can provide practical solutions and improve efficiency in traditionally manual fields – even if physical labor is still required.

The discussion turns to the future of work, with a focus on industries like transportation. The Liebermans stress that professions facing disruption have time to adapt if they proactively embrace innovation and collaborate rather than passively consuming technological advancements; otherwise, their roles could be significantly altered 15 years down the line.

The speakers discuss how AI corporations can easily increase prices on their systems, potentially exploiting users' dependence. They experienced recently that models haven’t become cheaper, highlighting a concerning trend where access to these powerful tools could be controlled and priced at the discretion of a few companies. This creates a power imbalance with significant implications for productivity and accessibility.

If access to AI models is guaranteed and affordable, the speakers envision a future where individual productivity increases fourfold. This surge in output could lead to a societal shift where people work less or adapt to increased abundance, potentially resulting in larger homes, smarter appliances (like self-buying refrigerators), and automated infrastructure like self-repairing pipes.

Drawing parallels to the addictive nature of social media, the speakers express concern that AI could exacerbate issues related to information consumption and decision-making. They warn that controlled sources of information can shape perceptions of reality, potentially eroding individual freedom if users are unable to critically evaluate AI-generated content or advice.

The speakers emphasize that successful AI labs now rely on attracting and retaining brilliant AI researchers, not just engineers, as engineering tasks are increasingly automated by AI itself. They note a trend of researchers leaving organizations due to ethical disagreements, suggesting that the influence of these individuals will grow while consumer power diminishes over time, increasing dependence on specific models.

The discussion highlights concerns about centralized control of AI data, specifically referencing 'World ID' as a black box controlled by a single company. This lack of transparency raises anxieties among users and creates a need for decentralized solutions where access isn't limited to one entity. The speakers emphasize the importance of open-source approaches that demonstrate data security and prevent monopolistic control.

The speakers express optimism about AI’s ability to foster equality by providing equal access to technology. They envision a future where individuals facing job displacement can leverage AI tools for guidance and new opportunities, such as starting businesses or learning new skills. This democratization of resources could unlock the highest potential for everyone, regardless of background.

A key takeaway is the necessity of embracing and experimenting with new AI tools. The anecdote about the artist who consistently adopted new technologies demonstrates how proactive learning can lead to rapid advancement. The speakers recommend dedicating time, even just an hour a day, to explore these tools rather than passively consuming entertainment.

A bold claim is made that AI will be capable of performing most intellectual work within three years. Despite this, the speakers assert that core human attributes like creativity, collaboration, and personal decision-making will remain essential. This suggests a shift in focus towards uniquely human skills rather than tasks susceptible to automation.