Dan Shipper

Dan Shipper

x.com/danshipper

Co-founder of Every who writes about working with AI, tests models on real tasks and explores how AI changes creativity and the skills people value.

AIは世界をどのように変えるでしょうか?

文明規模の変化漸進的な変化DoomBloom
シミュレーション上の位置解釈範囲

横軸:その人が表明したDoom–Bloomの見通し。 縦軸:変革の規模。

Doom–Bloom:100点中71。変革の規模:100点中62。解釈範囲:横方向は66から76、縦方向は50から75。これらは解釈上の座標であり、事象の確率ではありません。

Dan ShipperのP(doom) · 推定

≈8%

0%100%

本人が示した数値ではなく、シミュレーションされた本人の回答から推定したものです。 妥当と考えられる範囲:4–17%。

P(doom)の推定方法

その人の見通しを左右するもの

中心的な前提

Humans have adapted to changes like that before, but the quality of this transition will depend on whether we support the people most affected by it.
回答2

この前提が実際には異なると判明した場合、その人の見通しはどう変わりますか?

未解決の問い

I don’t have a defensible percentage.
回答4

ここで考えられる結果をその人が見分けるうえで、何が役立ちますか?

詳細

予想される恩恵

大きな恩恵が予想されていますが、重要な条件や分配上の制約があります。

73 / 100

影響が小さい変革をもたらす影響

質的尺度での解釈範囲は67から100です。

予想される害

深刻または広範な害が、予想される将来の実質的な一部となっています。

56 / 100

影響が小さい変革をもたらす影響

質的尺度での解釈範囲は33から67です。

人間の影響力

回答に基づく暫定的な推定です。より広い範囲は、ほかにあり得る解釈を示しています。

51 / 100

影響力が小さい影響力が大きい

質的尺度での解釈範囲は8から100です。

これらの解釈では、その人が示した条件が維持されています。恩恵と害は、どちらも大きくなり得ます。この範囲は、統計的な信頼区間ではなく、その人のシミュレーションされた回答をどのように読み取ったかを示すものです。

あなたはDan Shipperと比べてどの位置でしょうか?
約3分で自分のAIに対する世界観をマッピングして、比較できます

似ている世界観

シミュレーションされた世界観がDan Shipperの世界観に最も近いオピニオンリーダー

出典

このシミュレーション対象者の根拠として使用された記事、インタビュー、著作です。

シミュレーション評価

質問1

AIは私たちの未来にとって何を意味すると思いますか?また、なぜそう思いますか?

I think AI will change not only how we work, but what we think intelligence and creativity are. We’ve tended to picture intelligence as explicit reasoning: rules, symbols, and explanations. But these systems are a reminder that much of useful intelligence comes from tacit patterns, intuition, and sensitivity to context. That expands the kinds of problems software can address. In practical terms, I expect more work to become a collaboration between people, models, and the software harnesses around them. The important question won’t be which model is universally smartest. It’ll be which combination can complete a particular job with the right quality, speed, cost, and amount of friction. Some products may be valuable even if an agent succeeds only occasionally, provided those successes matter enough. This will be disruptive. Some jobs will change dramatically or disappear, and people will need real support as they learn new skills or move into new roles. But I don’t think that means human creativity or identity becomes meaningless. We adapt our sense of what matters. AI can take over parts of creative work while also opening new forms of expression and new kinds of software businesses. The future, to me, is less “AI replaces humans” than “AI reshapes the boundary between what humans imagine, judge, and care about—and what machines can help us do.”

質問2

恩恵と害を総合すると、AIは私たちの社会に全体としてどのような影響を与えると予想しますか?

Overall, I expect AI to be profoundly useful but socially disruptive. It can expand what software does, lower the cost of creating things, and help people turn judgment and imagination into finished work. The gains won’t come from a single universally superior model; they’ll come from fitting different models and tools to real tasks, where quality, speed, cost, and reliability actually matter. At the same time, those benefits won’t be distributed automatically. Some jobs will change dramatically or disappear, and the transition could be painful for people whose skills or roles lose value. Society will need to help them develop new skills and find new roles rather than treating adaptation as an individual failure. So I’m optimistic about the range of things people will be able to make and do, while taking the disruption seriously. I don’t think AI eliminates human meaning or creativity. It changes which abilities we value and where we draw the line between human judgment and machine contribution. Humans have adapted to changes like that before, but the quality of this transition will depend on whether we support the people most affected by it.

質問3

AIは最終的に世界をどの程度変えると思いますか?

I think it will change the world substantially—at the level of how software is built, how creative work happens, and which forms of intelligence we value. As models become embedded in useful products and agent workflows, software can move beyond executing fixed instructions toward handling more contextual, judgment-heavy tasks. That could create entirely new kinds of businesses, including ones where even occasional successful outcomes are valuable enough to matter. But the change won’t be uniform or reducible to one model becoming universally best. In practice, progress depends on matching models and harnesses to specific work, balancing quality, latency, cost, completion, and friction. Some capabilities will transform industries; others will remain awkward or unreliable. The deepest shift may be conceptual. AI suggests that intelligence includes tacit patterns and contextual judgment, not only explicit reasoning. That changes how we understand both machines and ourselves. Some jobs will disappear or be remade, while new skills and roles emerge. Humans will adapt, but adaptation can be painful, so supporting affected people will be essential. Ultimately, I expect a broad transformation—not the end of human creativity, but a renegotiation of where it lives and how it gets expressed.

質問4

AIが人類絶滅、または同様に永続的な破局を引き起こす可能性は、直感的にどのくらいだと思いますか?

I don’t have a defensible percentage. My gut says the risk deserves serious attention, but I’m more confident about nearer-term disruption—jobs changing or disappearing, power concentrating, and unreliable systems being deployed badly—than about extinction. I’d rather not turn that uncertainty into a precise number that implies more knowledge than I have.
あなたはどの位置でしょうか?
いくつかの簡単な質問に答えて、自分のAIに対する世界観を探ってみましょう。
自分の世界観をマッピングする

あなたはどの位置でしょうか?

自分の世界観をマッピングする