Scores by kind of claim
Each kind of claim is scored on its own, out of 100: the share that held up, with the range it likely falls in. Fewer resolved claims give a wider range.
-
Factual claims
Statements about the present or the past, checked against the record: did they hold up?
83 78 88 -
Predictions
Statements about events outside the person's control, checked once they fell due: did they come true?
Insufficient record: 2 predictions resolved, 5 needed for a score.
-
Promises
Commitments to the person's or organisation's own future action, with something to deliver: was it delivered?
Insufficient record: 0 promises resolved, 5 needed for a score.
All types combined: 83 (likely 78 to 88, high confidence), from 86 claims: 86 factual claims. It mixes kinds of claim in different proportions for different people, so compare people kind by kind.
Daniela Amodei's claims
38 claims
“People talk to an online therapist now, and sometimes that is backed by an AI system that is giving advice.”
Daniela Amodei, Anthropic president, interview 4 March 2022 Daniela and Dario Amodei on Anthropic (Future of Life Institute)
People use online therapists, and sometimes the advice is backed by an AI system.
Mostly true True in substance, wrong in a detail.
What would count as true or false
Judge as of 2022-03-04. True: evidence shows that people used online therapy services by then and that at least some such services used AI systems to generate or support the advice given to users (e.g., AI chatbot therapy apps such as Woebot or Wysa, or AI-assisted advice on therapy platforms). Mostly true: online therapy was widespread and AI was used in mental-health apps, but only in adjacent forms, such as self-help chatbots not presented as therapists or AI used for triage rather than advice. Mixed: AI's role was mainly administrative, with only marginal evidence of AI-backed advice. Mostly false: AI was involved only in non-advice functions such as scheduling or matching. False: evidence shows no online therapy or mental-health advice service used AI to give advice by that date.
Evidence
- Effectiveness and Safety of Using Chatbots to Improve Mental Health: Systematic Review and Meta-Analysis Journal of Medical Internet Research · 13 July 2020
“Chatbots in the remaining one-third of studies (4/12) utilized machine learning and natural language processing (artificial intelligence) to understand users’ replies and generate responses.”
- Is Online Therapy Right for Me? University of Utah Huntsman Mental Health Institute · 5 February 2021
“Although e-therapy has seen a significant spike in usage over the past nine months, some are still uncomfortable with the idea.”
- Wysa raises USD 20M in funding round - Medical Buyer Medical Buyer · 15 July 2022
“The firm provides an AI-enabled conversation platform to support users and in the second level provide access to high-level techniques with human support.”
Reasoning
E2 (February 2021) reports a significant spike in e-therapy usage, showing that people used online therapy before the claim date. E1 (a 2020 systematic review) shows that some mental-health chatbots used machine learning and natural language processing to understand users' replies and generate responses. However, the quote does not show these were presented as therapist advice on online therapy services rather than adjacent self-help chatbots. E3 describes Wysa's AI-enabled support platform but is dated after the judging date, so the evidence fits the mostly-true tier better than the full true tier.
Cite this claim
Daniela Amodei, “People use online therapists, and sometimes the advice is backed by an AI system.” (said 4 March 2022). Verdict: mostly true. Reliability API, claim c_c453e610af9d, method 0.1.0+28da144b. https://epistacklabs.com/claims/c_c453e610af9d
Claims without a verdict
Is a verdict here wrong, or do you speak for Daniela Amodei? Reply on the claim's own page, with the form at its foot (how replies work). Verdicts are made automatically; how verdicts and scores are made.