Apr 22, 2026 · 52m · big-technology
Are We Too Obsessed With AI Predictions? — With Carissa Véliz
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
Oxford philosopher Carissa Véliz joins host Alex Kantrowitz on the Big Technology Podcast to critically examine society's obsession with predictive AI, automated gatekeeping, and speculative prediction markets. Together, they explore the ethical and social costs of algorithmic forecasting while making a compelling case for preserving human agency and embracing an unpredictable future.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Alex holds 37.7% of the talking time here. How this is scored →
Alex as informed peer 5.4 Guest teaching 6.2 Guest disagreement 4.5 Alex pushing back 4.8
speaking balance: gold is Alex, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 7:27 Véliz rejects individual agency reframe as incentivizing fraud and stalking
Véliz rejects Kantrowitz's suggestion that applicants should bypass systems with direct outreach, arguing it rewards rule-breakers, fraudsters, and stalkers over genuinely qualified introverts.
Hardest push from Alex ▶ 36:09 Kantrowitz rejects Frankfurt 'bullshit' label for LLMsKantrowitz directly challenges Véliz's claim that LLMs are mere bullshitters, arguing that massive engineering investments in truth-grounding and proven coding utility demonstrate genuine economic value.
Biggest teaching moment ▶ 12:25 Véliz distinguishes actionable causal facts from opaque black-box patternsVéliz gives a clear pedagogical breakdown explaining why transparent bank criteria empower applicants to fix deficiencies, whereas black-box statistical predictions trap applicants in unverifiable, unfixable discrimination.
Alex holds their own ▶ 21:00 Kantrowitz demonstrates AI efficacy using Google's flood forecasting modelsKantrowitz counters Véliz's historical Oracle of Delphi comparison by detailing Google Research's flood prediction models, demonstrating how predictive machine learning tangibly saves lives in the real world.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Alex as informed peer | Guest teaching | Guest disagreement | Alex pushing back | Why |
|---|---|---|---|---|---|---|
| The Myth of Fixed Futures and Self-Fulfilling Prophecies | 5 | 6 | 3 | 2 | Kantrowitz introduces the topic of prediction mania and algorithmic steering. Véliz educates the host on how prediction creates self-fulfilling prophecies, comparing algorithmic elimination to a crime that leaves no error signal. | |
| Agency, Outlier Talent, and Algorithmic Gatekeeping in Hiring | 6 | 7 | 6 | 7 | Kantrowitz pushes back vigorously against Véliz's stance, arguing individuals have agency to circumvent automated hiring filters by emailing managers directly. Véliz firmly counters that this selects for rule-breaking and stalking behaviors while filtering out introverted or quirky geniuses. | |
| Financial Prediction, Contestability, and Algorithmic Lending Bias | 6 | 7 | 5 | 6 | Kantrowitz cites machine learning tools like C3 AI used by mortgage officers to defend algorithmic lending efficiency. Véliz refutes this by contrasting contestable causal facts with black-box correlations that perpetuate systemic racial discrimination. | |
| Counterfactual Limits and Kafkaesque Automated Bureaucracy | 6 | 6 | 4 | 5 | Kantrowitz suggests auditing algorithms for bias rather than discarding them entirely. Véliz explains the absence of counterfactuals and invokes Hannah Arendt to describe the psychological toxicity of Kafkaesque automated systems, which Kantrowitz concedes with a recent customer service example. | |
| Historical Divination, AI Interpretability, and Physical Prediction | 6 | 6 | 5 | 6 | When Kantrowitz questions AI interpretability and Véliz compares modern AI to the Oracle of Delphi, Kantrowitz pushes back by citing Google's life-saving flood prediction models. Véliz parries by noting Google's failed flu prediction attempts to distinguish physical models from social ones. | |
| Near-Term Forecasting, Mass Surveillance, and Democratic Freedom | 5 | 6 | 5 | 5 | Kantrowitz presses Véliz for concrete examples when she warns that predictive surveillance leads to a police state. Véliz presents crime data comparisons between Spain and the UK, and Kantrowitz brings up his firsthand observations of surveillance in Beijing. | |
| Probabilistic Distortions in the Criminal Justice System | 4 | 6 | 4 | 4 | Kantrowitz challenges the notion that society is nearing a 'Minority Report' reality. Véliz explains how probabilistic risk scores in bail, parole, and litigation insurance erode principled justice. | |
| The Anonymity Debate in Digital and Physical Spaces | 5 | 7 | 6 | 6 | Kantrowitz questions the necessity of anonymity in public protests, arguing it breeds online-style extremism in the physical world. Véliz rebuts his premise with empirical research showing identified figures online often drive the most aggressive behavior. | |
| Generative AI, Frankfurt's 'Bullshit', and Grounding in Truth | 7 | 6 | 6 | 7 | Véliz characterizes LLMs as sycophantic 'bullshitters' under Harry Frankfurt's definition because they lack truth-tracking design. Kantrowitz pushes back firmly, citing commercial economic value, tool calling, and grounding efforts, though Véliz illustrates base model reasoning failures with a logic puzzle. | |
| Mid-Show Break: Introducing the Prediction Markets Debate | 5 | 6 | 4 | 4 | Kantrowitz asks why prediction markets are surging in popularity and accuracy. Véliz outlines their vulnerabilities to market manipulation, conflict escalation, and insider trading, warning against gamifying geopolitical conflict. | |
| Humor, Art, and Reclaiming Unpredictability Against Algorithmic Determinism | 4 | 5 | 2 | 1 | Véliz explains why human humor and art—using the Seinfeld pilot as an archetype—serve as essential counters to predictive determinism. Kantrowitz agrees, noting that LLMs fundamentally struggle with comedic subversion. |
Statements from this episode (24)
Véliz: The most impactful personal, business, and societal events are least predictable
“The most important events in your life, in your personal life, but also in your business life and in our lives as a society are the ones that are the most unpredictable.”
Véliz: Predictive AI creates self-fulfilling prophecies with zero error signals
“Self-fulfilling prophecies are like the perfect crime. Because it's like a murder weapon that disappears upon striking. It leaves no record. It creates no error signals. We will never know how that person would have fared, because they will never get the job, …”
Véliz: Algorithmic hiring filters cause employers to lose key talent
“Society wants that talent. We're missing out on important talent when we streamline everything.”
Véliz: Academia faces a serious problem with data fraud
“In my world, in academia, I think we have a serious problem of fraud. Of people who are very well known, and who have been very successful, and who have fudged their data, or who have committed other kinds of academic fraud.”
Véliz: Predictive lending models are uncontestable and shroud injustice
“And I reject your application on the basis of a prediction. There's no way you can contest that, because predictions are not facts. At best, they're educated guesses. And because they're not facts, you cannot prove it to be false. And so, it's a way to shroud …”
Véliz: The Markup found mortgage algorithms denied Black applicants with identical profiles
“The markup had a very long story a few years ago about how two people who had applied for a mortgage had been denied, and when the markup investigated, their file looked exactly the same or very similar to other two people who happened to be white, and they, a…”
Véliz: Algorithmic bias audits are limited by a lack of counterfactual data
“When you say, well, let's investigate for bias or investigate for inaccuracy, there is a limit to what we can do, because we will never have the counterfactual. This is not a randomized control trial, right? And you still have the problem that without clear cr…”
Véliz: Opaque algorithmic systems induce alienation and magical thinking
“We are building systems that are very Kafkaesque, that are impossible to navigate, and I don't know if you've had this experience in which they are becoming so alienating and so Kafkaesque That people start having, like, magical thinking about the algorithm, a…”
Kantrowitz: Google's AI flood prediction models are accurate and save lives
“One of the things that they're really working hard on and Google research is
Flood prediction, which we know like kills way too many people because it's totally preventable.
that's not.
You know, do we know every single thing about how these machine learning …”
Véliz: Google Flu Trends failed because search query data was too noisy
“When Google tried to predict flus and pandemic type events it tried for years and years and years, it increased the complexity, it increased the data, and it could never do it, and eventually it shut down.
Partly because it was relying on people's searches,
Pe…”
Véliz: Mass surveillance leads to authoritarianism and a police state
“The price to pay for survey, for mass surveillance, is a police state. It leads to authoritarianism, and so often we're willing to surrender our privacy on promises that are never kept, that are very problematic, even if they could keep it, and we're sort of s…”
Véliz: The safest countries globally are not the most surveilled
“The safest countries in the world are not the most surveilled ones. So Spain is one example. It has some of the lower statistics for any kind of crime, including homicide and so on, violent crimes. And it's not better surveilled Than the US or the UK. And in f…”
Véliz: Facial recognition surveillance erodes protest anonymity vital to democracy
“When you have a protest, and in particular, a peaceful protest, It's very important to have anonymity. That is one of the bedrocks of democracy, and when you have cameras all over the place, and now with facial recognition being so easy to use, you are eroding…”
Véliz: Surveillance machinery exists primarily to feed predictive AI systems
“Surveillance is important because the whole machinery of surveillance is there to feed the machinery of prediction. So these two machineries are intimately related, and that's why it matters.”
Véliz: Predictive algorithms are actively used for sentencing in justice systems
“No, but we're using predictive algorithms in the justice system for sentencing, for many aspects in the justice system, and for the reasons that we explored with insurance, or with loans, or with jobs, that's very problematic.”
Véliz: Probabilistic justice allows bad actors to evade accountability easily
“When we make the justice system about probabilities, we're losing its principled approach. And so you make it very easy for the bad guys to get away with it, because you don't have to make it impossible for people to challenge you, or even very hard. You just …”
Véliz: People posting under real identities online are actually more aggressive
“Even though it's very intuitive to think that way, when you look at the empirical data, it shows that people who are identified online tend to be more aggressive, and then they tend to be more followed and more successful in that aggression.”
Véliz: Large language models have no regard for truth
“And that's essentially what a large language model is. It has no regard for the truth. It wants to please you. If what pleases you happens to be true, great. But if it's not true, then it doesn't care one way or another.”
Véliz: HBR study shows AI makes workers less productive
“There was a paper recently at the Harvard Business Review that suggested that even when people think they're being more productive with AI, when you have researchers look at it, they're being less productive because they're spending a lot of time fixing what t…”
Véliz: Politicians Bet on Themselves to Manipulate Prediction Markets
“If you want to influence public perception and you have enough money, you can bet heavily on something or someone to make it look more popular, and we already have examples of politicians betting on themselves.”
Véliz: Anonymous Wallets Made $1.2M Betting on Iran Attack
“Six anonymous accounts earned 1.2 million dollars on a prediction market betting for the attack on Iran, and some of those wallets were funded hours before, which suggests that they might have had insider information, and if they had insider information, did t…”
Véliz: Prediction Markets Provide No Productive Capital Unlike Stock Markets
“The stock market, when you invest in a company, you're actually contributing capital to that company in a way that is an important contribution to society. Whereas the prediction market is just a bet.”