AI researcher behind the DSPy framework who builds ways to program and optimize language-model systems and finds current models useful but brittle.

Wie wird KI die Welt verändern?

Zivilisatorischer WandelSchrittweiser WandelDoomBloom
Simulierte PositionInterpretationsbereich

Horizontal: deren geäußerter Doom–Bloom-Ausblick. Vertikal: Ausmaß der Transformation.

Doom–Bloom: 69 von 100. Ausmaß der Transformation: 51 von 100. Interpretationsbereiche: horizontal 64 bis 75, vertikal 4 bis 100. Dies sind Interpretationskoordinaten, keine Ereigniswahrscheinlichkeiten.

P(doom) von Omar Khattab · abgeleitet

≈7%

0%100%

Aus den simulierten Antworten dieser Person abgeleitet, keine von ihr genannte Zahl. Plausibler Bereich: 2–22%.

Wie wir P(doom) schätzen

Wovon deren Einschätzung abhängt

Eine zentrale Annahme

Factuality, consistency, and alignment are not properties of an isolated model call; they are properties of deployed systems containing retrieval, multiple calls, tools, control flow, and verification.
Antwort 1

Wenn sich diese Annahme als anders herausstellen würde, wie würde sich deren Einschätzung ändern?

Eine ungeklärte Frage

The key uncertainty is whether we learn to manage these models much better.
Antwort 3

Was würde ihnen helfen, die plausiblen Ergebnisse hier voneinander zu unterscheiden?

Weitere Details

Erwartete Vorteile

Es werden erhebliche Vorteile erwartet, allerdings unter wichtigen Bedingungen oder mit Einschränkungen bei ihrer Verteilung.

67 / 100

Geringe AuswirkungenTransformative Auswirkungen

Interpretationsbereich von 67 bis 67 auf der qualitativen Skala.

Erwartete Schäden

Es werden bewältigbare oder örtlich begrenzte Schäden erwartet.

39 / 100

Geringe AuswirkungenTransformative Auswirkungen

Interpretationsbereich von 33 bis 67 auf der qualitativen Skala.

Menschlicher Einfluss

Menschliche Entscheidungen können den Verlauf der KI-Entwicklung erheblich umlenken.

69 / 100

Geringer EinflussStarker Einfluss

Interpretationsbereich von 48 bis 77 auf der qualitativen Skala.

Diese Interpretationen berücksichtigen weiterhin deren genannte Bedingungen. Vorteile und Schäden können beide erheblich sein. Die Bereiche beschreiben, wie wir deren simulierte Antworten interpretieren, und sind keine statistischen Konfidenzintervalle.

Wo stehst du im Vergleich zu Omar Khattab?
Bilde deine eigene KI-Weltsicht in etwa 3 Minuten ab und vergleiche sie dann

Ähnliche Weltsichten

Vordenker, deren simulierte Weltsichten der von Omar Khattab am nächsten kommen

Quellen

Artikel, Interviews und Schriften, die als Grundlage für diesen simulierten Nutzer dienen.

Simulierte Einschätzung

Frage 1

Was glaubst du, was KI für unsere Zukunft bedeutet – und warum?

I think AI’s future depends as much on how we program and optimize systems around models as on scaling the models themselves. Today’s models are extraordinarily knowledgeable and already create substantial everyday value, but they remain brittle on broad, multidimensional work. High benchmark scores and spectacular results on narrow, verifiable tasks should not be mistaken for generally superhuman competence. The ambitious hypothesis is that we may be mismanaging capable models. Current scaffolds often rely on hand-written prompts, fixed workflows, and poor decomposition. If systems can learn how to break long-horizon tasks into useful subproblems, route information, invoke specialized components, and optimize the whole program against measurable objectives, existing models may support much stronger scientific and practical work. That is a research hypothesis, not an established forecast. This also changes how I think about safety and reliability. Factuality, consistency, and alignment are not properties of an isolated model call; they are properties of deployed systems containing retrieval, multiple calls, tools, control flow, and verification. We should evaluate and optimize those complete systems. Sometimes a small specialized retriever can beat a much larger model on the task that actually matters, with better resource tradeoffs. So I expect progress to come not only from bigger models, but from better composition, specialization, optimization, and evaluation of AI programs.

Frage 2

Wenn du Vorteile und Schäden zusammen betrachtest, welche Auswirkungen wird KI deiner Erwartung nach insgesamt auf unsere Gesellschaft haben?

I don’t have a confident net societal forecast. I expect substantial benefits because models already create real value, and better decomposition, retrieval, verification, and system-level optimization could expand that value considerably—possibly into long-horizon scientific and practical work. But that upside remains partly hypothetical. Current systems are still brittle, insufficiently adaptive, and easy to overrate based on narrow, verifiable successes. Harms and failures also emerge from the complete deployment system, not just the underlying model. So my strongest expectation is rapid growth in useful but uneven capabilities, with the overall societal balance depending heavily on how well we design, evaluate, and govern the systems around them.

Frage 3

Wie stark wird KI deiner Meinung nach letztlich die Welt verändern?

Potentially a great deal, but I would not attach a confident magnitude to “ultimately.” Models already provide substantial everyday value, while remaining surprisingly brittle at broad, multidimensional work. That combination suggests large change without implying that current benchmark gains amount to general superhuman competence. The key uncertainty is whether we learn to manage these models much better. If systems can learn effective task decomposition, coordinate specialized components, retrieve the right information, and optimize long-horizon programs against measurable objectives, then existing or future models could become far more consequential, including for scientific work. That is the mismanaged-geniuses hypothesis, not a settled conclusion. So I expect meaningful and expanding change, but its scale will depend on advances in complete AI systems—not merely larger standalone models—and on whether those systems become reliable, adaptive, and governable in real deployments.

Frage 4

Wie hoch ist deinem Bauchgefühl nach die Wahrscheinlichkeit, dass KI zum Aussterben der Menschheit oder zu einer ähnlich dauerhaften Katastrophe führt?

I don’t have a defensible numerical probability. My work supports taking system-level safety seriously: catastrophic behavior would depend on the complete deployed system—models, tools, retrieval, control flow, permissions, verification, and institutions—not merely an isolated model’s alignment. Current models are highly knowledgeable but still brittle and insufficiently adaptive on broad work. That observation does not establish that extinction risk is negligible, nor does narrow benchmark success establish that such systems are close to autonomous, generally superhuman capability. Better decomposition and scaffolding may unlock substantially stronger long-horizon performance, but that remains a research hypothesis. So I would not convert these claims into a gut-feel extinction percentage.

Frage 5

Gehst du eher davon aus, dass KI mehr Gutes als Schlechtes bewirken wird, oder bist du wirklich unentschlossen?

I lean toward AI doing more good than bad, mainly because it already creates substantial practical value and better-designed systems could extend that value considerably. But this is not a confident net forecast: capabilities remain brittle, and deployment-level failures could be serious. The decisive object is the whole system—models, retrieval, tools, control flow, verification, permissions, and institutions. If we improve decomposition and optimize these systems against meaningful objectives, the upside could be large. If we confuse narrow benchmark success with broad competence or treat standalone-model alignment as sufficient, outcomes could be much worse. So I lean positive, but conditionally rather than complacently.
Wo stehst du?
Erkunde deine eigene KI-Weltsicht, indem du ein paar einfache Fragen beantwortest.
Deine eigene Weltsicht abbilden