Author David Baldacci Testifies on AI Copyright Infringement at Senate Judiciary Hearing

Summary:

Author David Baldacci testified before the Senate Judiciary Committee about the impact of AI on authors and copyright. He highlighted:

His shock upon discovering AI models like ChatGPT were trained on his pirated novels, allowing them to generate stories in his style, effectively stealing his creative work.
The distinction between human learning (reading for inspiration) and AI training (ingesting vast amounts of copyrighted material for profit without permission).
His frustration that trillion-dollar AI companies claimed licensing individual works was too difficult, resorting to using pirated content instead.
The devastating consequences for the publishing industry, including market saturation with cheap AI-generated content, reduced profits for publishers, and decreased investment in emerging human writers.
His argument that if AI's use of copyrighted works is considered "fair use," then no one, including AI companies themselves, has copyright protection.
His belief that books, created through human effort and imagination, remain more truly transformative than AI.

Introduction to the Hearing and David Baldacci's Background [0:00]

David Baldacci, a bestselling author, was introduced at the Senate Judiciary Committee hearing. [0:00]
- The hearing's focus was on the "AI industry's mass ingestion of copyrighted works." [0:00]
Baldacci drew a parallel to Mark Twain's 1905 appearance before Congress to advocate for stronger copyright laws. [0:27]
- Twain, the most pirated author of his day, believed creative arts were the nation's lifeblood. [0:33]
Baldacci shared his personal journey as a writer. [1:11]
- Reading books in the library helped him travel the world and develop a desire to write. [1:14]
- He worked for decades, enduring rejections, honing his craft, and remaining disciplined. [1:23]
- After publishing 60 novels, he continues to work hard, believing in the "American way" of perseverance. [1:33]

Baldacci's son prompted ChatGPT to write a plot in David Baldacci's style. [1:42]
- Within five seconds, three pages were generated containing elements from Baldacci's previous books. [1:49]
- This included plot lines, twists, character names, and narrative styles. [1:53]
He then discovered that the AI community had taken most of his novels without permission. [1:57]
- His works were fed into machine learning systems. [2:01]
- Baldacci described feeling as if someone "backed up a truck to my imagination and stolen everything I'd ever created." [2:06]

Baldacci directly refuted the argument that AI's actions are comparable to an aspiring writer reading books to learn. [2:09]
- He stated this comparison is "flatly wrong." [2:20]
He used his own experience with John Irving's novels as an example. [2:23]
- Although he read everything Irving wrote, his own novels do not read like Irving's. [2:27]
- Unlike AI, a human cannot remember every line, detail, character, or plot point. [2:34]
Humans read other writers for appreciation and motivation, not to copy or steal. [2:41]
AI, conversely, uses writers' output as an "incredibly valuable shortcut" and "super fuel" to teach its software programs. [2:50]

Baldacci highlighted that trillion-dollar AI companies did not purchase his books; they obtained them from pirated websites. [2:58]
- These companies claimed it would be too difficult to license works from individual creators. [3:05]
- He called this an "efficient" way to steal. [3:09]
He questioned the legitimacy of this excuse. [3:11]
- Trillion-dollar companies with "battalions of lawyers" should have the resources to operate lawfully. [3:11]
- As a former trial lawyer, he stated such an argument would be dismissed in court. [3:17]
He argued that if AI companies only needed words, they could have used dictionaries. [3:24]
- Instead, they specifically needed "complete, well-crafted, living, breathing stories" with realistic characters, logical plots, and genuine dialogue. [3:40]
- They relied on the human "craft... learned with the sweat of our brows and the flexing of our imaginations." [3:51]
These companies "swooped in, stolen that labor in order to make enormous profits," while writers receive nothing. [3:55]

AI enables anyone to generate a novel "in the vein of an established writer" with no effort. [4:05]
- Such books can be sold, claiming they read like a specific author's work, e.g., David Baldacci's. [4:09]
- Baldacci asserted, "Yes, it does read like my novels because it is my novel. It is my imagination." [4:13]
He compared the situation to "cheap imported goods hurting American workers," but in this case, it's "cheap books being created by American technology flooding the market." [4:21]
- This will lead to lower profits for publishers. [4:28]
- Less money will be available to invest in new and emerging writers, harming the entire industry. [4:30]
Online vendors are already requiring authors to disclose if a book was not human-created. [4:35]
- Baldacci believes it will soon be necessary to limit the number of books someone can publish daily or weekly, which he considers "insane." [4:41]

Baldacci pointed out that source code and algorithms are also protected by copyright. [4:48]
- He hypothesized that if he stole AI source codes and profited, he would face numerous lawsuits. [4:52]
He challenged AI's "fair use" argument regarding his body of work. [5:00]
- If his entire work is considered fair use, then "there is no more copyright protection for anyone." [5:05]
While AI expects its own intellectual property to be fully protected, it does not extend the same courtesy to authors. [5:09]
He concluded by contrasting AI's purported "transformation" with the long-standing transformative power of books. [5:24]
- Books have transformed billions of people, shaped human history, and emerged humanity from the Dark Ages. [5:27]
- Books teach empathy and make the world "kinder, gentler, more meaningful." [5:41]
- He firmly believes that books will always be more transformational than AI. [5:58]