Hacker News Digest
Monday, October 12, 2026
In This Issue
- Hacker News
- Passkeys are a trap. Do not use them
- Total Annihilation, rebuilt for modern computers
- Lobbying is corruption
- Andrej Karpathy breaks down the capability gap
- NYC's "Click to Cancel" rule is now in effect
- Fingerprints Are the New Cookies
- Fable Took My Job
- Artificial Insanity
- "Robot" Is a Social Construct: Why Your Dishwasher Is Not a Robot
Zipper Data Brief
October 12, 2026
Your daily digest of the best from Hacker News
Top 6 Trending
#1
698 points
· laurex
· comments
# Summary
The discussion reveals a fundamental divide: while passkeys offer superior security against phishing and MITM attacks compared to passwords/TOTP, critics worry that attestation mechanisms could enable vendor lock-in, restrict software freedom, and force device upgrades—concerns that advocates argue are hypothetical and shouldn't override passkeys' current security benefits.
#2
550 points
· jttnr
· comments
# Summary
The discussion centers on modern remakes of classic RTS games, particularly highlighting **Beyond All Reason** as the best spiritual successor to Total Annihilation, with mentions of other active communities around Supreme Commander and Forged Alliance Forever. Commentary also touches on broader themes of IP preservation, AI-generated content detection, and nostalgia for the original games' design and soundtracks.
#3
521 points
· LucidLynx
· comments
# Summary
The discussion revolves around whether lobbying itself is inherently corrupt or a necessary democratic function. Commenters split between those arguing that transparent, expertise-based lobbying is essential for informed policymaking, and those contending that the money involved and revolving-door politics make it effectively legalized corruption.
#4
503 points
· albertzeyer
· comments
# Summary
The discussion reveals a sharp divide in AI capability understanding: a tiny elite (0.2%) has access to frontier models and agent swarms delivering transformative productivity gains, while billions remain unaware or underwhelmed by current public tools. The gap persists because most people lack both the technical skills to leverage these tools and tangible use cases, though this will gradually close as companies build domain-specific products on top of frontier models.
#5
498 points
· gregsadetsky
· comments
# Summary
NYC's new "Click to Cancel" rule requires subscription services to offer easy cancellation matching signup convenience, though commenters note enforcement concerns and worry companies may circumvent it with early termination fees or longer contract commitments. Multiple states already have similar laws, and the EU is expected to implement comparable rules soon.
#6
440 points
· royfish
· comments
# Summary
Fingerprinting has become a pervasive alternative to cookies for tracking users online, exploiting browser and hardware data that's difficult to block. The discussion reveals that while technical defenses exist, the real solution requires legislation—particularly EU-style regulations treating tracking as a privacy violation requiring explicit consent.
AI / Machine Learning
177 points
· jllyhill
· comments
# Summary
Software engineers face genuine job displacement due to AI capabilities, but the threat is driven by management's short-term profit motives rather than AI's actual ability to fully replace human judgment and domain expertise. The discussion reflects anxiety about technological disruption, with commenters debating whether AI will truly eliminate software roles or merely change them, while acknowledging that workers shouldn't over-identify with their jobs given capital's historical pattern of replacing labor when economically convenient.
61 points
· leglock
· comments
# Summary
The discussion is fundamentally divided: proponents argue LLMs have demonstrably moved beyond "stochastic parrots" by solving hard mathematical problems and showing genuine capabilities, while skeptics contend they lack true intelligence, continuous learning, and consciousness, and that progress is being extrapolated unrealistically from recent gains that will eventually plateau. Both sides accuse the other of moving goalposts and missing the real issues around AI's societal impact.
50 points
· birdculture
· comments
# Summary
The discussion explores how "robot" lacks a strict definition and is instead a social construct—what qualifies as a robot depends on context, language, and cultural norms rather than objective criteria. Commenters illustrate this by noting that similar devices are classified differently across fields (French food processors called robots, traffic lights in South Africa, CNC machines), and that once machines become reliable everyday tools, we stop calling them robots.
24 points
· nobody9999
· comments
# Summary
The discussion centers on skepticism about Anthropic's claims that LLMs like Claude could suffer or feel pain, with commenters debating whether this is genuine concern or marketing hype, while noting the fundamental uncertainty about machine consciousness remains unresolved despite strong opinions on both sides.
27 points
· droidjj
· comments
The discussion highlights concerns that AI-generated legal filings are becoming bloated, incoherent, and could force judges to rely on AI tools themselves, potentially ending human-centered jurisprudence and creating an arms race where both parties use AI to overwhelm the other with meaningless content.
Startups / Business
12 points
· devonnull
· comments
# Summary
The commenter suggests that IPOs function as mechanisms for insiders to offload inflated valuations onto retail investors while limiting their own liability.
All startups need is distribution
9 points
· relaxfolio
· comments
# Summary
Distribution is startups' biggest challenge—most technical founders lack sales skills and can't effectively reach customers, making advertising and marketing expertise essential but often overlooked.
114 points
· m463
· comments
# Summary
The discussion is divided between skeptics citing New Constructs' poor IPO prediction track record and concerns about Anthropic's profitability/valuation, and optimists arguing Anthropic has a strong competitive moat in enterprise AI despite high costs and open-source competition.
4 points
· mraniki
· comments
No comments available.
3 points
· colors__
· comments
The commenter expresses cautious interest in the concept but is skeptical about its scalability, suggesting it would likely only attract a small user base before losing momentum.
More Stories (34)
398 points
· gumby
· comments
227 points
· bhrlady
· comments
147 points
· doppp
· comments
141 points
· cainxinth
· comments
207 points
· rooftopzen
· comments
95 points
· andsoitis
· comments
80 points
· MilnerRoute
· comments
64 points
· robaato
· comments
97 points
· espeed
· comments
69 points
· hxii
· comments
431 points
· heyimada
· comments
64 points
· ayaros
· comments
69 points
· Vaslo
· comments
Ask HN: Do you want a router that has DD-WRT pre-installed?
66 points
· Ixander
· comments
83 points
· ljf
· comments
68 points
· eustoria
· comments
144 points
· luggage_bazooka
· comments
425 points
· Anon84
· comments
42 points
· hn_acker
· comments
41 points
· DeepLogin
· comments
27 points
· jacquesm
· comments
45 points
· smurda
· comments
28 points
· zdw
· comments
36 points
· ishanz
· comments
20 points
· MilnerRoute
· comments
141 points
· lukaslalinsky
· comments
70 points
· andsoitis
· comments
122 points
· saisrirampur
· comments
79 points
· optimalsolver
· comments
27 points
· MilnerRoute
· comments
17 points
· fullstacking
· comments
20 points
· LouisLazaris
· comments
17 points
· sbulaev
· comments
178 points
· amosbatto
· comments
Created by Zipper Data Co.
· 2026-10-12 12:02 UTC
· Unsubscribe
Get digests like this delivered to your inbox every morning.
Subscribe Free