Amicus Presents: The Pre-Crime Unit

Podcast: Amicus With Dahlia Lithwick
Featured Podcast: Hi-Phi Nation (Host: Barry Lam)
Date: February 9, 2019
Episode Focus: The rise of algorithmic, predictive policing in the LAPD, its philosophical implications, and community pushback.

Overview

This episode, curated by Amicus, presents the first installment from Slate’s “Hi-Phi Nation,” hosted by Barry Lam. “The Pre-Crime Unit” probes the realities of predictive policing—law enforcement’s use of data and algorithms to forecast and prevent crime. Using Los Angeles as a case study, Lam explores the legal, societal, and ethical ramifications of this new technology, centering on community responses, how these tools work, and the profound questions they raise about fairness, privacy, and justice.

Key Discussion Points & Insights

1. CSOC & Algorithmic Policing: Unfolding Debate in LA

• LAPD Board Meeting On CSOC Funding:
  • Police commission reviews a $35K grant for a Community Safety Operations Center (CSOC)—now LAPD's hub for predictive policing.
  • Community members, particularly the Stop LAPD Spying Coalition, object, citing surveillance and racial targeting.
  • Memorable Moment: Public comment section is tense and repeatedly interrupted, displaying deep mistrust.
    • "[CSOCs] are the central nerve centers... to criminalize members of the community." – Stop LAPD Spying Member [03:26]
    • Vote passes despite protests.

2. Predictive Policing Today versus Sci-Fi Precedents

• Minority Report vs. PredPol:
  • Fictional “pre-crime” inspired by Minority Report has partially materialized, but with twists:
    • Real tools forecast property crime, drug offenses, and gun violence—affecting marginalized neighborhoods.
    • Predictive policing isn't about free will or psychics: it's about statistical inference from vast datasets.
    • “In the real world, free will isn’t the issue. The real philosophical problems are more basic. But maybe even harder.” – Barry Lam [08:49]

3. How PredPol Works (Place-Based Prediction)

• Guest Expert: Dr. Sarah Brayne (UT Austin sociologist)
  • PredPol: Analyzes historical property crime (location, time, type); outputs “predictive boxes” officers patrol during uncommitted time.
    • “It’s not that different than what they were doing before…” – Sarah Brayne [12:43]
  • Critique: Risk of “over-policing” the same neighborhoods due to feedback loops, as data comes from police sources.
  • Defenses: Proponents point to race-agnostic, published algorithms akin to earthquake forecasting.

4. The Next Generation: Social Data & Real-time Prediction

• Research Highlight: Flora Salim (RMIT University)
  • Uses crowdsourced Foursquare “check-in” data to predict crime within three hours—with up to 16% better accuracy than PredPol. [17:51]
  • The more diverse, less regular the crowd, the greater the crime prediction.
    • “We managed to have improvement of up to 16%.” – Flora Salim [17:58]
  • Raises concerns about privacy, especially as technologies (e.g., Fitbit) could further monitor both suspects and police stress.

5. Operation LASER: Person-Based Predictive Policing

• Operation LASER:
  • Uses “chronic offender” bulletins generated via algorithms—targeting individuals predicted most likely to offend.
  • Data largely stems from “FI cards” (field interview notes) detailing any police-civilian interaction, including bystanders.
    • “Even people that have never talked to the police and have no direct police contact are included in law enforcement databases.” – Sarah Brayne [27:30]
  • System can generate self-perpetuating surveillance: Police target those with high scores, document further encounters, which build scores further.
    • “…that might turn into somewhat of a self-fulfilling prophecy…where if you’re going out and specifically seeking out the people with high points value…that can very quickly lead to a feedback loop.” – Sarah Brayne [29:12]

6. Community Impacts and Civil Liberties Concerns

• Stop LAPD Spying Coalition’s Critique:
  • Strong evidence that racialized communities—especially Black and brown youth—are over-represented and harmed.
  • Predictive algorithms and associated police discretion create “race-neutral” tools with disparate impact.
    • “We found in 2017, the Black community was five times more often arrested than the white community…data can stand in as a proxy for race.” – Jamie Garcia [33:07]
  • Civil rights issues: Extraction from neighborhoods, stigmatization via nuisance abatement and housing eviction, chilling effects on social mobility and trust in institutions.

7. Philosophical and Legal Dilemmas

• Statistical Evidence vs. Individual Justice:
  • Guest: Renee Bollinger (Philosopher, Australian National University)
    • Warns that “error costs”—the danger of wrongfully targeting innocents—are high even with strong statistics.
      • “…even for probabilities that clear a threshold for reasonable belief…if the error costs are high enough, you shouldn’t treat it as true.” – Renee Bollinger [38:36]
    • The more frequently a generalization is applied, the greater risk of systemic injustice and eroded trust—especially for marginalized groups.
      • “It’s possible for a statistical generalization to be both accurate and unjust. Sometimes the more accurate it is, the more unjust it can be.” – Barry Lam [45:15]
  • Raises the specter of changing standards of “individualized suspicion” in law—will big data undercut probable cause and individualized rights?

8. Larger Societal Implications

• Risk Exposure & Surveillance:
  • Routine encounters with police—even absent arrests—cause affected individuals (disproportionately minorities) to avoid banks, schools, and hospitals, reducing upward mobility and trust in institutions.
    • “…we really need to think about not just what the benefits for increased police efficacy can be, but what the chilling effects of surveillance can be, too.” – Sarah Brayne [46:44]

Notable Quotes & Memorable Moments

• “Algorithmic objectivity is a fiction, a cover.” – Narrator, summarizing coalition views [05:13]
• “You’re rubber stamping CSOCs and you’re rubber stamping the same goddamn policies.” – Stop LAPD Spying Coalition Member [06:23]
• “Palantir can then give you a social network map for an individual, who in the past they’ve been seen with, cars they’ve driven, where in the neighborhood they’ve been stopped at…” – Narrator [27:15]
• “The most consequential question is whether it’s going to lead to a revolutionary change in the standards of reasonable suspicion and probable cause…” – Barry Lam [36:10]
• “It’s possible for a statistical generalization to be both accurate and unjust. Sometimes the more accurate it is, the more unjust it can be.” – Barry Lam [45:15]

Timestamps for Major Segments

| Time | Segment | |------------|------------------------------------------------------------| | 01:09 | Introduction to LAPD Police Commission/CSOC | | 03:06 | Tense public comment period | | 07:38 | Minority Report & predictive policing parallels | | 11:04 | Interview: Dr. Sarah Brayne - Explaining PredPol | | 14:51 | Interview: Dr. Flora Salim - Social data & real-time AI | | 22:46 | Hi-Phi Nation bonus content, transition to Operation LASER | | 24:23 | Stop LAPD Spying Coalition confronts LAPD on Operation LASER| | 25:58 | The mechanics of FI cards and chronic offender scoring | | 29:12 | The feedback loop in predictive targeting | | 33:07 | Disproportionate impact (race as a proxy) | | 38:06 | Philosophical critique: statistics vs. individualized suspicion | | 43:39 | Paradox of statistical discrimination & risk exposure | | 45:41 | Chilling effects on communities |

Tone & Final Thoughts

The episode is at once analytical and urgent—grounded in philosophy but alive with street-level activism and high-stakes debate. It invites listeners to grapple with complex realities: sophisticated predictive tools may promise safety, but at significant costs to civil rights, trust, and equity—especially for the most marginalized.

The discourse is candid, with activists expressing deep anger and academics offering nuanced, sometimes paradoxical insights into how data-driven policing may change the very standards of legal justification and justice.

For further exploration: The next episode promises to trace predictive algorithms from the streets into the courtroom and prisons, questioning how big data will reshape American notions of guilt, innocence, and due process.