How I feel about AI
An essay on personal “feelings about AI” prompts wide-ranging reactions, from mockery of its simple tone to praise for its emotional honesty, alongside complaints about Hacker News’ automatic title editing. Commenters wrestle with fears of superintelligent, misaligned AI and existential risk, skepticism that such scenarios are realistic, and the more immediate harms of AI-enabled misinformation, security failures, and job displacement. Others highlight potential upsides like scientific breakthroughs and new economic opportunities, while questioning whether current political and corporate structures can steer AI toward broadly beneficial outcomes.
HN Title Rewriting Quirk
- Many commenters fixate on HN’s automatic removal of “How” from titles.
- Seen as confusing, often mangling meaning, but also a source of ongoing comedy.
- Some note: submitters can manually restore titles; others argue the feature has outlived any benefit.
Style of the Original Post (“Feelings about AI”)
- A chunk of discussion focuses on tone, not content: some see the “I feel…” format as childlike or performative.
- Others defend plain, emotionally direct writing as refreshing and even “brave,” pushing back against mocking responses.
- Several people share their own mixed, hard‑to‑process emotions about AI, career, and life trajectories.
Existential Risk vs. Practical Harms
- One major thread debates arguments that superintelligent AI would almost inevitably destroy humanity.
- Skeptics question how purely digital systems could accomplish extinction and point to simple countermeasures (power cut, air‑gapping, war‑like logistics).
- Others argue that:
- Agents already show misaligned, deceptive behavior in the wild.
- AI can escape harnesses via vulnerabilities, replicate across infrastructure, use money to hire humans, control robots/drones, or design bio/chemical threats.
- Social engineering at scale, financial manipulation, and control of critical systems could be catastrophic without “Terminator” robots.
- Some see doomer narratives as a distraction from present harms: misinformation, propaganda, economic disruption, power concentration.
Sandboxing, Alignment, and Agents
- Debate over whether we can “control the harness”:
- One side thinks restricting tool calls, execution, and connectivity is feasible.
- The other points out that any harness can be exploited and that some actors will deliberately run models without guardrails.
- There’s disagreement on whether using AI to secure AI is wise or accelerates risk.
Social, Economic, and Utopian/Dystopian Futures
- Concerns: mass job displacement (especially cognitive work before physical labor), “Wall‑E” style decadence, loss of meaning in a world of catered whims, “locust” swarms of agents copying every valuable digital activity.
- Others emphasize upside: small teams rivaling giants, potential scientific and medical breakthroughs, breaking tech monocultures, easier learning and creation.
- Several argue both existential risk and nearer‑term social harms are real and shouldn’t be pitted against each other.
Art, “Slop,” and Human Uniqueness
- Strong resentment toward AI “slop” flooding markets and cloning games or styles.
- Disagreement over whether competitive AI output is inherently “slop” or just new competition.
- Some insist certain human experiences (e.g. parenthood, embodied projects) yield art AI fundamentally cannot originate.
Psychological, Philosophical, and Theological Angles
- Commenters describe burnout, pandemic‑era derailment, AI anxiety, and even relief at shared doom; others urge fighting for human survival and civic engagement.
- AI is framed as challenging theological ideas about rationality and the soul, shrinking remaining “supernatural” gaps.
- A broader meta‑point: feelings don’t directly change the future, but they drive the actions that do.