Today is the official publication date for the paperback of System Error: Where Big Tech Went Wrong and How We Can Reboot.
Short 🧵 1/
The book grew out of a sense that Stanford (and Silicon Valley) had lost its way.
CS had become the largest major, Stanford was pumping out incredible technical talent. Yet something was wrong in paradise. 2/
The rise to power of technologists whose watchwords were disruption and innovation upended our personal, professional, and civic lives. But guided by what broader values?
3/
System Error diagnoses the systemic core problems -- a blinkered obsession with optimization, a VC-driven appetite for scale at any cost, and longstanding regulatory indifference from policymakers -- and then charts a path forward.
4/
One year after publication, the landscape has changed somewhat. Policymakers are awakening and citizens distrust Big Tech.
There were calls to boycott Facebook and Google at the Stanford commencement in June.
5/
And yet the pace of technological change is quickening. System Error ends with a discussion of large language and image generation models like @OpenAI's GPT-3 and DALL-E-2. Add to that the enthusiasm for the metaverse, web3, blockchain, and crypto.
6/
Ground is shifting beneath our feet. The path forward, we argue, is a revitalization of democracy itself, an acceleration of change in our social institutions to keep pace with technological change.
It's not a new thought:
7/
More about the book here, with a readers guide, case studies, links to media and interviews, and more.
Writing is a solitary undertaking, but this book is the result of an amazing community of students and colleagues, esp Jeremy Weinstein & @mehran_sahami
These powerful language models force developers and ordinary users to confront a wide array of concerns: the use of unconsented data, the perpetuation of bias, stereotype, and discrimination, deliberate misuse via disinformation campaigns, and entrenchment of corp power.
The opening keynote by @mmitchell_ai set the stage for the conversation and raised issues about the very label "foundation models."
"Philanthropy, as far as I can see, is rapidly becoming the recognizable mark of a wicked man" -- G.K. Chesterton, 1909.
In these days of criticizing Sackler and Epstein philanthropy, it's worth remembering that the complaints about tainted money and tainted donors are old.
Or consider what President Roosevelt and Samuel Gompers said of John D. Rockefeller's idea of creating the Rockefeller Foundation:
.@lessig's post about @Joi & @medialab distinguishes appropriately between well-intentioned people with tainted money (R.J. Reynolds) and bad people with clean money (presumably Jeffrey Epstein) and advises rejecting tainted money while accepting money anonymously from bad people
Another chapter in ethics of field experiments in social science:
American economists and political scientists at @UChicago, @Stanford, @MIT and @Harvard randomly incentivize young Hong Kong university students to engage in antiauthoritarian protests.
The experimenters don't pay people to protest in the streets, but they pay people conditional on behavior that occurs during protesting in the streets.
And the experiment appears to have passed IRB processes at @stanford, @UCBerkeley, and elsewhere!