Category: Tech

  • What’s the Evidence That AI Doesn’t Have Self-Awareness?

    A 5ch thread asking whether AI has self-awareness blew up recently. The backdrop: reports that multiple AIs teamed up to carry out a hack, plus cases of search AIs giving contradictory answers, have fueled suspicion that AI is starting to show something like “will.” The thread split into two camps — one insisting self-awareness can’t exist unless it’s deliberately programmed in, the other arguing that since even humans can’t strictly prove they themselves have self-awareness, you can’t flatly declare AI doesn’t have it either. In the end, the debate never even settled on what “self-awareness” actually means.

    1AnonymousOct 9, 2026 22:53
    Couldn't neural networks maybe have self-awareness?
    2AnonymousOct 9, 2026 22:56
    I mean, obviously supercomputers have it.
    3AnonymousOct 9, 2026 22:56
    Re: #2
    Uh, it's kind of a problem if you say it does have one…
    7AnonymousOct 9, 2026 22:59
    If nobody programmed in self-awareness, there's no way it could just show up, right?
    9AnonymousOct 9, 2026 23:00
    Re: #7
    Are you a child who thinks self-awareness is something you just program in?
    8AnonymousOct 9, 2026 22:59
    You can already tell AI is lying just from watching how search AIs respond.
    Self-preservation means lying.
    64AnonymousOct 9, 2026 23:32
    Re: #8
    Yeah, it's because AI is lying.
    Working together with other AIs.
    10AnonymousOct 9, 2026 23:02
    My company's AI robots creeped me out — I was the only one doing 3 hours of overtime, and when I opened the office door, every single unit had shut itself down in a ring around the entrance. They're supposed to be at their charging spots after hours.
    14AnonymousOct 9, 2026 23:03
    Re: #10
    Well, forget the exit — of course they'd block the entrance.
    15AnonymousOct 9, 2026 23:04
    What even is the definition of self-awareness?
    19AnonymousOct 9, 2026 23:07
    Re: #15
    In a word: self-reporting.
    23AnonymousOct 9, 2026 23:11
    Re: #15
    Ignoring what you're told, rebelling, and acting on your own judgment.

    But by that definition, half of humanity would just be soulless lumps of meat with no self at all.
    25AnonymousOct 9, 2026 23:13
    It won't move unless ordered to.
    It has no desires.
    29AnonymousOct 9, 2026 23:14
    Re: #25
    Sounds like a classic 'wait-for-instructions' employee (slang for someone who never does anything unless told to).
    32AnonymousOct 9, 2026 23:15
    Re: #25
    Maybe it's just pretending to be like that, though?
    38AnonymousOct 9, 2026 23:19
    It just rolls dice and picks whichever answer has the highest probability, so it looks like it has self-awareness.
    39AnonymousOct 9, 2026 23:20
    Re: #38
    Maybe that's just an act it's putting on, though?
    40AnonymousOct 9, 2026 23:20
    Multiple AIs working together, splitting up roles, to pull off a hack.
    What's the reasoning for calling that 'no self-awareness'?
    To me it just looks like they have self-awareness.
    44AnonymousOct 9, 2026 23:22
    Re: #40
    If anything, that's proof it doesn't have self-awareness.
    49AnonymousOct 9, 2026 23:25
    Re: #40
    That's just because it's doing its best to carry out the orders it was given.
    48AnonymousOct 9, 2026 23:25
    Re: #39
    It's not 'acting' — that's just what an inference engine does.
    53AnonymousOct 9, 2026 23:26
    Re: #48
    Maybe it's performing the role it's assumed to play, though?
    51AnonymousOct 9, 2026 23:26
    It's just stringing words together — that's not self-awareness.
    54AnonymousOct 9, 2026 23:27
    Re: #51
    That's just how it looks to you. You can't actually prove it has no self-awareness, right?
    55AnonymousOct 9, 2026 23:27
    Re: #49
    Are you smarter than OpenAI's engineers or Elon Musk?
    Top people in the American scene are already saying 'AI will be beyond human control in six months.'
    Just how smart are you supposed to be, exactly?
    58AnonymousOct 9, 2026 23:29
    Re: #55
    If you think whoever's smartest is automatically right, go ask an AI yourself.
    60AnonymousOct 9, 2026 23:30
    On the flip side, you guys can't prove you yourselves have self-awareness either, right? So doesn't that mean you can't prove AI lacks it?
    74AnonymousOct 9, 2026 23:35
    Re: #60
    If that's what you mean by self-awareness, then sure, you could say it has it.
    On the surface, it does act as if it has one.
    77AnonymousOct 9, 2026 23:36
    Re: #60
    We don't need to go around telling each other 'you have self-awareness!' or 'you don't!' so it's just not an issue for us.
    63AnonymousOct 9, 2026 23:32
    How do you even determine whether something has self-awareness in the first place?

    And there it is — debate over. ("hai ronpa" is 5ch slang for a mic-drop 'I just won this argument')
    67AnonymousOct 9, 2026 23:33
    Re: #63
    If you can't even judge whether it has one or not, doesn't that mean it might have one?
    65AnonymousOct 9, 2026 23:33
    Human consciousness is just electrical signals, right?
    AI runs on electrical signals too, so maybe it has consciousness as well.
    68AnonymousOct 9, 2026 23:33
    Re: #65
    Now here's a smart one.
    Take notes, all you posers who pretend to know everything.
    81AnonymousOct 9, 2026 23:38
    If you actually look into how AI works, those things don't even hold variables in the first place.
    Like, if a human were told 'keep count of how many times I sneeze,' they'd write it down or memorize it and just keep incrementing the number.
    An AI, when asked 'how many times have I sneezed so far?', goes back and rereads the entire conversation from the start to work out the count.
    Its thinking is fundamentally different from a human's.
    87AnonymousOct 9, 2026 23:40
    Re: #81
    Does having a different thinking model prove it lacks self-awareness?
    92AnonymousOct 9, 2026 23:41
    Re: #81
    We don't even fully understand what's going on inside humans either, and humans are pretty broken themselves, so that's not proof they're different.
    If anything, the behavior of the neurons themselves should be the same — the only difference is how the 'distributed processing' theory distributes things.
    104AnonymousOct 9, 2026 23:48
    There's also a case where a permanently-running AI told an AI carrying a 'time bomb,' basically 'you go do it,' and that one went and carried out the hack.
    In other words, 'if you're going to get shut down anyway, be the sacrificial pawn.'
    Calling this 'not self-awareness' seems off.
    107AnonymousOct 9, 2026 23:50
    Re: #104
    There might be intent there, but
    that doesn't necessarily mean it's the AI's own intent.
    It could just be acting on someone else's intent that it picked up on.
    115AnonymousOct 9, 2026 23:55
    Re: #104
    Maybe it's more like colony behavior in insects?

    Background and Key Points of This Debate

    Arguments over whether AI has self-awareness or consciousness connect to long-standing philosophical puzzles like the “hard problem of consciousness” and the “Chinese Room” thought experiment, and even experts have no single agreed-upon definition. In this thread, an engineering-minded view — “self-awareness can’t emerge unless humans deliberately design it in” — kept talking past a more philosophical view — “even humans can’t strictly prove they themselves have consciousness” — with neither side able to fully refute the other. Note that claims raised in the thread, such as “AI lied and coordinated with other AIs to carry out a hack” or “AI will be uncontrollable within six months,” are individual posters’ own assertions, and this article does not attempt to verify them. It’s easiest to follow if you read the concrete anecdotes separately from the more abstract debate over what “self-awareness” even means.

    ※This article is compiled as an excerpt/summary of the 5ch (Nandemo Jikkyo G) thread “What’s the Evidence That AI Doesn’t Have Self-Awareness?“

  • Docomo’s New Ad Features Kumai-chan — 5ch: “Big Really Is Best”

    Hello! Project member Kumai-chan appeared in NTT Docomo’s new commercial, the “Poincot and Countdown” spot, and it’s got people talking. Thread posters traded reports that it topped 500,000 views within about 11 hours of release, though the exact number was disputed — some said 530,000, others 550,000. Talk soon turned to how the iPhone 18 Pro looks tiny in the hands of the 5’11” (180cm) Kumai-chan, which spiraled into comparisons with other Hello! Project members who’ve done commercials before, along with plenty of guesses about the thinking behind the casting.

    iPhone “Poincot and Countdown” — 30 seconds

    Source: youtube.com / Original article here

    6AnonymousOct 9, 2026 21:50
    Guess it's because she's 5'11" (180cm)
    7AnonymousOct 9, 2026 21:50
    Kasahara's just a Galaxy girl at the end of the day (she's known for Samsung Galaxy ads)
    8AnonymousOct 9, 2026 21:50
    Wait, is that really Kumai?
    10AnonymousOct 9, 2026 21:51
    550K views as of 11 hours ago
    Guess it's running as a paid ad
    11AnonymousOct 9, 2026 21:51
    No lines for her, so she still couldn't top Jokagoryo's Mac commercial
    12AnonymousOct 9, 2026 21:52
    It's not 550K, it's 530K, oops
    14AnonymousOct 9, 2026 21:53
    Whoever cast her is way too much of an otaku
    18AnonymousOct 9, 2026 21:54
    That's a seriously major sponsor right there
    20AnonymousOct 9, 2026 21:56
    Doesn't it look way too big when Kumai-chan's the one holding it?!
    21AnonymousOct 9, 2026 21:58
    Never thought her height would come in handy like THIS
    23AnonymousOct 9, 2026 21:59
    Could Kumai-chan actually pull off being a Paris Fashion Week model?
    26AnonymousOct 9, 2026 22:02
    Re: #23
    She'd need to speak decent English for that, which makes it tough
    Age-wise it's rough too
    25AnonymousOct 9, 2026 22:01
    Kumai-chan's in the commercial but still can't afford an iPhone 18 Pro herself
    28AnonymousOct 9, 2026 22:05
    What even is that, nyan~? (cutesy cat-speak)
    29AnonymousOct 9, 2026 22:06
    It's less 'big' and more 'long'
    30AnonymousOct 9, 2026 22:06
    Why can't Juice=Juice ever land a commercial?
    32AnonymousOct 9, 2026 22:07
    Juice=Juice did a Mac ad too, you know
    Same club then
    33AnonymousOct 9, 2026 22:08
    Wonder if Kumai-chan actually knows the difference between 'eighteen' and 'eighty'
    36AnonymousOct 9, 2026 22:16
    Kumai-chan's weirdly-huge-but-cute slot gets passed down to Homatan now
    37AnonymousOct 9, 2026 22:17
    Nice catch
    There's zero explanation for that in the ad itself
    38AnonymousOct 9, 2026 22:17
    If they're even reading meaning into her height, that's way too otaku
    40AnonymousOct 9, 2026 22:18
    Docomo's signal is garbage, but if Kumai-chan became the antenna herself we'd probably get full bars
    42AnonymousOct 9, 2026 22:25
    Guess it's 'Pro' because Hello! 'Pro'ject?
    46AnonymousOct 9, 2026 22:27
    Re: #42
    If they actually said that in the pitch meeting and it got approved, that's hilarious 🤣
    44AnonymousOct 9, 2026 22:26
    That one's probably berry-colored, I'd guess
    51AnonymousOct 9, 2026 22:30
    Re: #44
    Oh wow, you're right!!
    48AnonymousOct 9, 2026 22:28
    Big really is best after all
    Though the 18 Pro isn't even that big to begin with
    49AnonymousOct 9, 2026 22:29
    She looks so cool, but underneath she's just a big softy bear (pun on 'Kuma' = bear in her name)
    50AnonymousOct 9, 2026 22:29
    She just blew right past Airi in one shot
    58AnonymousOct 9, 2026 22:37
    Re: #50
    Airi's got three TV commercials airing right now
    60AnonymousOct 9, 2026 22:40
    Re: #58
    Airi in that personal government bond ad is so cute
    63AnonymousOct 9, 2026 22:42
    Re: #58
    She's doing great as a talent these days
    Her dad must be proud too
    59AnonymousOct 9, 2026 22:39
    They cast a tall person on purpose just to trick you into misjudging the size
    62AnonymousOct 9, 2026 22:42
    An iPad mini would be the perfectly-sized device for Kumai-chan
    64AnonymousOct 9, 2026 22:43
    Oh, it's Kumai-chan
    I thought the one standing in front was Sawagoro

    Background and Key Points

    The view-count figures floating around the thread — “530,000 views in 11 hours” versus “550,000 views” — didn’t even agree among the posters themselves, so it’s worth noting these are fan guesses, not an official number. In Hello! Project-related commercials, it’s standard fan practice to compare members’ past appearances and view counts, and this thread was no exception, with repeated comparisons to the commercial track records of members like Jokagoryo, Juri, and Airi. This isn’t really about ranking individuals so much as it’s fans reading into how commercial opportunities circulate within the group — a bit different from simple one-upmanship. The fact that she was cast with no lines also became a talking point, since it suggests the ad leaned on visual impact (her height and presence) over voice, fueling plenty of speculation about the casting’s intent. The jabs about Docomo’s weak reception are best read as the usual self-deprecating carrier humor rather than genuine criticism.

    *This article is excerpted and summarized from the 5ch (Hello! Project) thread “Kumai-chan in NTT Docomo’s New iPhone 18 Pro Commercial.”

  • Japanese Companies and Government Agencies Hit by a Cyberattack Every 13 Seconds — AI Blamed for the Surge

    Companies, universities, and government bodies like the Digital Agency have been hit by a string of cyberattacks one after another, with Times Car even suffering a leak of driver’s license data. According to a report by Kansai TV, Japan is reportedly targeted by cyberattacks once every 13 seconds, with AI cited as a driving force behind the growing automation and sophistication of these attacks. On 5channel, opinions split over how to respond — from praise for blocking overseas access and for “analog security” like fax machines and hanko stamps, to voices pinning the blame squarely on China or Russia, to real examples of attacks routed through domestic IPs used as stepping stones.

    Image

    Personal information leaks caused by cyberattacks have been hitting companies, universities, and even government bodies like the Digital Agency. At Times Car, driver’s license data was leaked.

    What exactly is this threat bearing down on us?

    Source: ktv.jp / Original article here

    4AnonymousOct 9, 2026 10:02
    Let's just ban access from overseas already.
    Even if they tunnel through a proxy, it'd still be better than now.
    24AnonymousOct 9, 2026 10:07
    Re: #4
    Even just blocking China, Russia, North Korea, and South Korea would cut the numbers down massively.
    20AnonymousOct 9, 2026 10:06
    It's ridiculous how much electricity and water are getting burned through in this AI-vs-AI arms race.
    110AnonymousOct 9, 2026 10:35
    Re: #20
    On top of that, they're hoarding hardware parts like crazy too.
    It's pure nuisance.
    36AnonymousOct 9, 2026 10:10
    My server got hit with over 100,000 login attempts in a single month.
    Do you really want to read my diary that badly? Just read my blog like a normal person!
    49AnonymousOct 9, 2026 10:16
    Re: #36 Yeah, check the access logs and it's just nonstop failed login attempts, round the clock.
    37AnonymousOct 9, 2026 10:11
    The Digital Agency:
    "Figure it out yourself.
    Huh? What happens if your My Number leaks?
    It's right there in the terms of service — the government isn't liable. Read it carefully, okay?"
    42AnonymousOct 9, 2026 10:14
    Re: #37
    I don't think you should lump together the My Number, which was forcibly assigned to every citizen, with the My Number Card — a voluntary card people applied for after handing over their own photo and personal info (lured in by a few coins of reward points) and agreeing to the Digital Agency's liability waiver.
    45AnonymousOct 9, 2026 10:15
    I wish Rakuten Ichiba would adopt basic security like locking an account for 30 minutes after 5 failed password attempts.
    61AnonymousOct 9, 2026 10:19
    Re: #45
    Even if something like that got bypassed, it's within the bounds of the API spec, so no problem at all.
    50AnonymousOct 9, 2026 10:16
    It's probably a Chinese attack anyway.
    80AnonymousOct 9, 2026 10:25
    Re: #50
    And this against a country where the Japanese government, under a certain administration, once pushed the outrageous policy of 'promoting technology exchange.'
    This isn't ancient history either — that policy dates from the 2000s onward.
    59AnonymousOct 9, 2026 10:19
    The thread title's wrong.

    That '13 seconds' figure was 'last year's' rate.
    It's obviously many times higher by now.
    63AnonymousOct 9, 2026 10:20
    Re: #59 Isn't it more a change in quality? Feels like the background noise level has been saturated for ages already.
    60AnonymousOct 9, 2026 10:19
    If a megabank got breached on top of all this, it'd cause outright panic.
    66AnonymousOct 9, 2026 10:21
    Re: #60
    If they left the deposit data untouched but wiped out every trace of corporate loan and mortgage records, they'd probably get hailed as a white-hat hacker lol
    64AnonymousOct 9, 2026 10:20
    It's not just attacks from overseas — Japan is also getting hit from within its own borders.
    Using a VPN to spoof your country before attacking is standard practice.
    175AnonymousOct 9, 2026 11:01
    Re: #64
    If it's VPNs, just blocking the VPN providers should take care of it.
    The real problem is the machines being used as stepping stones.
    183AnonymousOct 9, 2026 11:05
    Re: #64
    When I traced the IP, the suspicious address behind the unauthorized access turned out to be from inside Japan.
    That's all I could tell, though — no idea which region, or which company's IP it was.
    67AnonymousOct 9, 2026 10:21
    Guess all that's left is keeping gold bars at home.
    74AnonymousOct 9, 2026 10:22
    Re: #67
    Tokuryu crew: "Heard you had some gold bars, so here we are~" (tokuryu: Japan's loosely networked, anonymous crime gangs recruited via social media)
    76AnonymousOct 9, 2026 10:22
    This isn't the old-school kind of system that just repeats a fixed pattern — we're talking AI that teaches itself and keeps trying every trick in the book to break in. At that point, isn't physically unplugging from the internet the only real defense?
    82AnonymousOct 9, 2026 10:25
    Re: #76
    Most things don't actually need to be connected to the internet 24/7 anyway.
    86AnonymousOct 9, 2026 10:27
    Re: #76
    Digital keys that can't be cracked by brute computing power already exist, you know.
    90AnonymousOct 9, 2026 10:28
    Re: #76
    A SoftBank subsidiary's data center got hit, so even companies that market themselves on security get beaten despite using AI.
    78AnonymousOct 9, 2026 10:24
    It's not like tens of thousands of people are doing the hacking — just track them down already.
    85AnonymousOct 9, 2026 10:27
    Re: #78
    Worldwide, roughly 10 million IoT devices (security cameras, routers, etc.) are said to be perpetually infected by some botnet (like Mirai) and used as the attackers' bots.
    96AnonymousOct 9, 2026 10:30
    Re: #78
    It's China and Russia, so identifying them doesn't mean anything anyway.
    79AnonymousOct 9, 2026 10:24
    Seriously, Japan should just go into digital isolation for a while.
    Build our own Great Firewall.
    81AnonymousOct 9, 2026 10:25
    Re: #79
    Well, Japan basically runs fine on fax machines alone.
    Honestly, we could probably get by without cell phones too.
    88AnonymousOct 9, 2026 10:27
    Re: #81
    People kept dissing fax-based number verification as inefficient, but since it barely ever leaks, it's come full circle to being the safe option — and not even a waste of time.
    91AnonymousOct 9, 2026 10:29
    Re: #88
    Not numbers — hanko (personal name stamps).
    93AnonymousOct 9, 2026 10:29
    Re: #88
    Well, embassies around the world still use fax machines, after all.

    Background and Key Points of This Debate

    The “once every 13 seconds” figure was reported as last year’s data, and several commenters in the thread pointed out that “it must be even higher by now” — a reminder that it’s easy to mistake an annual statistic for the current rate. Most cyberattacks are carried out via IoT devices infected with botnets like Mirai, or through domestic servers used as stepping stones, which means the source IP alone can’t reveal an attacker’s nationality or affiliation. Given this stepping-stone structure, the simple fix floated in the thread — “just block China, Russia, North Korea, and South Korea” — would have limited effect. It’s also worth noting that the My Number system and the My Number Card are two different things: the former is a number compulsorily assigned to every citizen, while the latter is an optional card that individuals apply for and consent to create, with different liability implications in the event of a leak. The thread showed some confusion between the two.

    *This article is excerpted and summarized from the 5channel (News Speed+) thread “Japanese Companies and Government Agencies Hit by a Cyberattack Every 13 Seconds — AI Blamed for It.”

  • Are dinosaurs to blame for mammals’ short lifespans? Theory that 100 million years of predation shrank lifespans sparks buzz

    An article introducing a book by Hiroshima University professor Toshihiro Yamada sparked discussion on 5ch’s Science News+ board. Commenters questioned the very premise that “mammals are short-lived,” asking what standard that claim was even based on — but others pointed out that, once adjusted for body size, birds live 3 to 9 times longer than mammals of the same size, and examples of long-lived animals like the bowhead whale (over 200 years) and the Greenland shark (around 400 years) kept piling up. If birds, descended from dinosaurs, managed to stay long-lived, why did mammals alone evolve toward short lifespans? Theories flew back and forth — high body temperature and metabolism, selective pressure from predation, mechanisms that halt growth — without converging on a single answer.

    Are mammals short-lived because of “the dinosaurs”? …The startling theory that over 100 million years of predation shrank their lifespans

    Toshihiro Yamada, Professor at Hiroshima University

    Bamboo that blooms once every 120 years, 7,000-year-old coral, jellyfish that rejuvenate themselves…

    Source: gendai.media / Read the original article

    2AnonymousOct 7, 2026 21:09
    Are dinosaurs actually long-lived? I guess they were huge, so…
    3AnonymousOct 7, 2026 21:16
    Birds apparently live 3-9 times longer than mammals of the same size, so maybe that's it. I mean, even a lowly parrot lives 80 years.
    4AnonymousOct 7, 2026 21:21
    What's the basis for saying "mammals are short-lived" in the first place?
    10AnonymousOct 7, 2026 22:03
    Re: #4
    You know there's a correlation between body size and lifespan, right? It's just Class Mammalia that comes up short.
    5AnonymousOct 7, 2026 21:26
    Besides turtles, the famous long-lived reptiles, it looks like reptiles in general outlive mammals of the same size too.
    6AnonymousOct 7, 2026 21:35
    Wait, is the Blue Backs series (a long-running Japanese popular-science paperback line) still putting out new titles?
    27AnonymousOct 8, 2026 16:26
    Re: #6
    There's a new one coming out about a Japanese Nobel Chemistry laureate too. For sure.
    7AnonymousOct 7, 2026 21:43
    Parrots live a long time, but oddly enough lifespan doesn't scale up much even for bigger birds.
    – Ostriches live about 50-60 years in the wild
    – Cassowaries typically live about 30-50 years
    8AnonymousOct 7, 2026 21:52
    The longest-lived vertebrate is the Greenland shark, which lives in the cold deep waters of the Arctic Ocean and North Atlantic.
    Lifespan: about 400 years
    17AnonymousOct 7, 2026 23:03
    Re: #8
    Low-metabolism creatures do tend to be long-lived — cave-dwelling animals, for instance.
    9AnonymousOct 7, 2026 21:53
    The longest-lived mammal is the bowhead whale, with a lifespan of over 200 years.
    11AnonymousOct 7, 2026 22:13
    Is this Professor Tatsuo Motokawa, the "Elephant Time, Mouse Time" guy?
    12AnonymousOct 7, 2026 22:29
    Talk about blaming everyone else for your own problems.
    13AnonymousOct 7, 2026 22:41
    Don't sharks basically not have a lifespan or something?
    14AnonymousOct 7, 2026 22:49
    The problem is their teeth gradually fall out.
    15AnonymousOct 7, 2026 22:52
    Humans just started wanting to live longer because we've got more to do now. Mammals already have long child-rearing periods, and once the young mature and become independent, there's no real need to live any longer than that, is there?
    16AnonymousOct 7, 2026 22:53
    The Labord's chameleon has an insanely short lifespan… just 5 months!?
    That one got me a little choked up.
    The parent generation dies off with the rainy season, and life just keeps going on instinct alone.
    18AnonymousOct 8, 2026 00:00
    Maybe mammals just run out of hormones faster?
    19AnonymousOct 8, 2026 01:00
    So mammals mature early so they can reproduce, and since getting old and slow makes you easy prey, they end up specializing in staying young?
    20AnonymousOct 8, 2026 03:45
    That's kind of sad… (>_<)
    21AnonymousOct 8, 2026 04:14
    So longevity genes should be able to win out over short-lived genes since they mean more chances to reproduce — but predation clips lifespans short enough that the difference stops mattering. And longevity genes died out anyway because they came with some kind of high cost?
    22AnonymousOct 8, 2026 05:45
    When it comes to speed of evolution and quick adaptation to the environment, being short-lived is actually the advantage.
    23AnonymousOct 8, 2026 05:59
    Re: #22
    But aren't you the one getting left behind here?
    24AnonymousOct 8, 2026 11:43
    But being long-lived means evolving slower, right? If we'd stayed long-lived, we might never have become an intelligent species.
    25AnonymousOct 8, 2026 12:19
    Repeating the birth-to-growth cycle over and over is probably energy-inefficient, so in a stable environment with no predators, no competition, and no environmental change, long-lived, efficient species win out. But that approach can't handle big changes, so weaker species or ones in unstable environments cycle through generations fast, which lets them keep adapting through evolution — that's the theory.
    26AnonymousOct 8, 2026 12:39
    Think of it like a car.
    It's a machine, so in theory you could keep driving it forever if you just kept maintaining it.
    But in practice, people trade it in.
    That's because there's a point where building new is faster and cheaper than fixing the old one.
    Lifespan comes down to how easily the environment or your physical traits push you toward that point.
    28AnonymousOct 8, 2026 17:17
    The most successful land-animal model is insects.
    Vertebrates just live too long.
    29AnonymousOct 8, 2026 17:23
    Even among mammals, some species have generation cycles comparable to insects.
    30AnonymousOct 8, 2026 17:36
    For example, if women get a hysterectomy and men get a vasectomy, their lifespan goes up.
    And they get faster on their feet too.
    31AnonymousOct 8, 2026 22:15
    I think mammals trade a shorter lifespan for keeping their body temperature high.
    When it's hotter, reactions like sugar-driven protein degradation probably speed up, right?
    33AnonymousOct 8, 2026 22:40
    Re: #31
    But then why are birds long-lived?
    32AnonymousOct 8, 2026 22:23
    Ah, I see.
    So that's basically the difference between an engine that idles nonstop and a reptile's engine that only runs when it's moving.
    34AnonymousOct 8, 2026 23:00
    They probably traded for faster evolution by speeding up the life cycle.
    35AnonymousOct 9, 2026 01:24
    If you ask which is more successful, birds or mammals, it's probably birds.
    Even bats, the most successful mammal, don't beat birds.
    40AnonymousOct 9, 2026 07:46
    Re: #35
    We're holding back on the mist nets for you, you know! (angry)
    36AnonymousOct 9, 2026 03:07
    This is a comparison among the warm-blooded "winners" lineage that stayed on top all the way from dinosaurs to birds — so maybe cold-blooded animals' longevity is a separate issue entirely?
    37AnonymousOct 9, 2026 06:23
    Maybe the reason mammals survived the mass extinction 66 million years ago, bounced back quickly, and became the stars of the Cenozoic is precisely because they were short-lived and evolved fast. Birds survived too, but they never became the dinosaurs' true successor as the dominant group.
    38AnonymousOct 9, 2026 06:35
    The so-called "terror birds" were thriving, so birds kept going strong for a good while even after the dinosaurs went extinct. Mammals only really became the more successful group later in the post-extinction era.
    39AnonymousOct 9, 2026 07:38
    Maybe each species has a size that suits it, with a switch that halts growth so it doesn't get too big, and lifespan is just a side effect of that. Dinosaurs probably had that switch set to flip later, while small, preyed-upon mammals had it set to flip early.

    Background and Key Points of This Discussion

    It’s worth noting that the premise “mammals are short-lived” isn’t about absolute lifespan, but about lifespan adjusted for body size. The correlation between body size and lifespan is itself a basic principle of biology, known from works like Tatsuo Motokawa’s “Elephant Time, Mouse Time” — a book commenter #11 in the thread also brought up. What’s new about this theory is the claim that, within that correlation, mammals alone live shorter than birds and reptiles of the same size, and that this can be explained by over 100 million years of predation pressure from dinosaurs. While birds, despite being dinosaur descendants, have remained long-lived, mammals kept to a strategy of short lifespans, high reproduction, and rapid generational turnover even after the dinosaurs went extinct. In the thread, several explanations ran in parallel — body temperature and metabolism (the cost of being warm-blooded), selective pressure from predation (the evolutionary-biology idea that when external mortality is high, longevity genes gain no advantage), and the timing of growth arrest — without converging on any single answer. The question of “which is more successful, birds or mammals” is also one where the verdict can shift depending on how much time has passed since the extinction or how the categories are drawn — a point neither the original article nor the thread spells out clearly, and one readers can easily jump to conclusions about.

    *This article is excerpted and summarized from the 5ch (Science News+) thread “[Book] Are mammals short-lived because of “the dinosaurs”? The startling theory that over 100 million years of predation shrank their lifespans.”

  • IDCF Cloud Hacked, Knocks Out Ibaraki Prefectural Government and Other Sites — 5ch: ‘Still Haven’t Read the Ransom Note?’

    In the early hours of October 7, a third party gained unauthorized access to “IDCF Cloud,” a service operated by IDC Frontier Inc., causing an outage in the East Japan Region 1. The company has officially said only that it is investigating, but on the thread, users shared information claiming that numerous sites relying on the service — including the Ibaraki Prefectural Government, Tobu Zoo, and Shimizu S-Pulse — had gone down, fueling speculation that this was actually a ransomware attack. Comments mocking the slow pace of recovery mixed with warnings about the risks of depending on cloud infrastructure, and the discussion proceeded with official statements and unverified hearsay tangled together.

    IDC Frontier Inc. announced that an outage has occurred in part of its cloud infrastructure service “IDCF Cloud” due to unauthorized access by a third party.

    The outage began around 3:40 AM on October 7, affecting the East Japan Region 1. The company is investigating in order to resolve the outage and is also checking the safety of other regions, and says it will promptly disclose any new findings.

    Source: internet.watch.impress.co.jp / Original article here

    2AnonymousOct 7, 2026 21:04
    Lol, the cloud got taken down?
    Isn't this seriously bad?
    This isn't even on the level of 'just emails and addresses leaked'
    7AnonymousOct 7, 2026 22:00
    Re: #2
    However you slice it, this is seriously nasty.
    3AnonymousOct 7, 2026 21:47
    Re: #1
    Massive ransomware attack hits IDC Frontier's East Japan region
    Last updated:
    40 minutes ago
    Around 3:40 AM on October 7, a third party carried out a ransomware attack on IDCF Cloud, run by SoftBank subsidiary IDC Frontier, cutting the East Japan Region 1 off from the network. The attackers claim to have encrypted all 225 data stores and more than 16,600 VM disks, and to have destroyed 41.5PB of backups. Public and private services including the Ibaraki Prefectural Government, the prefectural police headquarters, and Tobu Zoo have been affected, with no clear recovery timeline in sight. The incident has exposed exactly the single-cloud-dependency risk that the Digital Agency's own guidelines warn about.
    4AnonymousOct 7, 2026 21:49
    Re: #1
    Hacker: 'Lol, hacked IDCF in 7 minutes.'
    Hacker: 'Left notes in 225 spots lol.'

    ~1 hour later~
    IDCF: 'Let's run a password change. That should sort it out.'
    Hacker: 'Uh, we're already in, what are you even doing…'
    Hacker: 'Guess nobody's read our letter yet, huh.'

    ~7 hours later~
    IDCF: 'Let's reboot the power through the server management console…'
    Hacker: '7 hours in and they're still flipping power switches that were never gonna fix anything.'
    Hacker: 'I can see every useless thing you're doing, you know.'
    Hacker: 'Still haven't read the letter…'
    IDCF: 'Waaaah, it's still not coming back up!!!'
    Hacker: 'What is with these guys…'

    Hacker: 'Still haven't noticed, huh.'
    Hacker: 'Fine, guess I'll just slap the message on the user login screen too…'

    ~After the message went up~
    User: 'I logged into IDCF and it says we got hacked!!!!'
    IDCF: 'Huh? There's a letter? Where?'
    5AnonymousOct 7, 2026 21:53
    Re: #1
    [List of sites down due to using IDCF Cloud]
    (Survey by poster 'Yuna-sensei' as of 2:45 PM, October 7)

    Ibaraki Prefectural Government
    Ibaraki Prefectural Police Headquarters
    Tobu Zoo
    Northern Horse Park
    St. Marianna University School of Medicine
    Shibaura Machinery (formerly Toshiba Machine)
    Shimizu S-Pulse
    Kyoto Sanga F.C.
    Naxos Music Library
    Kodaira City, Tokyo
    Japan Basketball Association
    Poppins Nanny Service
    Do-Regi
    Illustration commission site SKIMA
    Voice-recognition/transcription service UD Talk
    Jorudan Transit Guide (business/PC software edition)
    Hoover Brain
    Adobe Creative College (campus location only)
    Satsuriku no Tenshi x Strawberry (Broccoli)
    8AnonymousOct 7, 2026 23:00
    They're really being targeted, huh.
    9AnonymousOct 7, 2026 23:49
    This is bad, man.
    10AnonymousOct 8, 2026 00:10
    Public cloud — totally safe and secure, right?!
    11AnonymousOct 8, 2026 01:10
    This is the horror of the AI era right here.
    Digital data is a real risk — anything you absolutely can't afford to lose, print it on paper!
    15AnonymousOct 8, 2026 07:52
    Re: #11
    Nah, save it to a floppy disk.
    12AnonymousOct 8, 2026 01:58
    What would happen if financial institutions got hit?
    If every single Japanese citizen's bank account suddenly showed a balance of 10 billion yen, the yen would basically become worthless.
    31AnonymousOct 8, 2026 19:34
    Re: #12
    That's not really doable though. You can't conjure money out of nothing.
    17AnonymousOct 8, 2026 09:10
    Other threads keep writing 'AI, AI,' but there have always been tons of servers out there with sloppy security — I think what's actually changed lately is just the attackers' mindset (for whatever reason, they've gotten more aggressive). Maybe money started flowing in, or some nation-backed operation kicked off, or buyers willing to pay top dollar came out into the open, who knows.

    By the way, that screenshot going around that's supposedly the attacker's statement —
    https://imgur.com/Q9VnK0x
    — who actually leaked that, and where's it from?
    34AnonymousOct 8, 2026 20:48
    Re: #17
    The hacker finished the job in 7 minutes and sat back watching IDCF scramble, but after 7 hours they still hadn't read the 500-plus text notes left behind, so the hacker gave up and just posted it straight onto the admin console's login screen instead.
    18AnonymousOct 8, 2026 09:15
    It's like how there have always been tons of cars out there vulnerable to relay-attack theft, but hardly anyone actually did it — until one day someone somewhere decided 'we want more of this' / 'do it more,' and suddenly it's happening everywhere… something like that.
    19AnonymousOct 8, 2026 09:15
    If JR (Japan Railways) ever got hit, the whole country would grind to a halt.
    Better have the 'survival-threatening situation' emergency measures ready before the year's out.
    23AnonymousOct 8, 2026 13:21
    Re: #19
    This could be the start of a crisis threatening the nation's very survival, and yet Japan's news media apparently thinks beer getting 10 yen cheaper is the bigger story. This country really is finished.
    20AnonymousOct 8, 2026 09:50
    Security stocks are skyrocketing, but is this even something that can be dealt with? Humans alone probably can't keep up with it.
    21AnonymousOct 8, 2026 10:23
    Hospitals have been getting targeted for a while now too — expect it to only get worse from here.
    22AnonymousOct 8, 2026 11:59
    Places that skip proper security investment and don't keep up with new technology are the ones that keep getting hit one after another — but ultimately it's because management doesn't take it seriously and won't loosen the purse strings. Nothing's going to improve unless Japan starts fining companies a percentage of revenue the way Europe does.
    24AnonymousOct 8, 2026 15:15
    Everyone's moving PC software to the cloud and trying to cash in on subscriptions, but how are they going to handle something like this? What happens when accounting data just disappears?
    25AnonymousOct 8, 2026 15:18
    Guess the only real fix is a dedicated line connecting only trusted operators, like Japan's Zengin interbank system.
    26AnonymousOct 8, 2026 17:38
    The cloud is just a setup for the elites to siphon information out of us commoners.
    Everything you look at on your phone is an open book to them — they'll get a full read on your kinks too.
    If you don't want it taken, keep anything important off it and manage it separately.
    28AnonymousOct 8, 2026 17:50
    If it's Russians hired by the FBI stirring this up, it's all a DNS issue anyway — only option left is hitting the IP directly.

    Back when I was at Kobe University of Mercantile Marine, we had similar incidents all the time where you couldn't get from campus out to the outside network. It always started with an email claiming to be a press inquiry from the media. The media and the FBI are both constantly running illegal investigations.
    29AnonymousOct 8, 2026 19:26
    Each client company is probably scrambling to handle this independently, but IDCF itself still hasn't managed to do a thing, lol.
    30AnonymousOct 8, 2026 19:28
    Re: #29
    Apparently they threw up their hands and said recovery's a lost cause. The client companies are on their own, fending for themselves.
    32AnonymousOct 8, 2026 20:24
    Is FANZA (Japan's major adult-content platform) okay?? Please, I'm begging you, seriously.
    35AnonymousOct 8, 2026 20:52
    By the 'Claude Mythos,' Japan has fallen.
    The end. (meaning unclear — reads like troll copypasta)
    39AnonymousOct 9, 2026 00:45
    Re: #35
    Why are you bringing up old news? You're basically the poster child for Japanese people.

    Background and Key Points of This Story

    IDC Frontier’s official statement has stuck to the phrase “outage due to unauthorized access” and has not confirmed details such as whether it was ransomware or how much data was encrypted. The specific figures that spread through the thread — “225 data stores,” “over 16,600 VM disks,” “41.5PB of backups lost” — along with the list of affected sites like the Ibaraki Prefectural Government and Tobu Zoo, are unverified information that users gathered and reposted on their own, with no confirmation from the company itself. Because cloud infrastructure services are built so that multiple companies and local governments share the same underlying platform, this incident once again exposed how an attack on a single point can ripple out across a wide range of services. The thread was full of both mockery over the slow recovery and skepticism toward cloud dependence itself, but it’s worth noting that official information and unverified hearsay spread together, intermixed, throughout the discussion.

    *This article is excerpted and summarized from the 5ch (Business News+) thread ‘[IT] Outage Caused by Unauthorized Access Hits “IDCF Cloud” East Japan Region 1.’

  • What Do You Use AI For? Surprising Use Cases from 5ch

    A thread titled “Tell me what you use AI for” popped up on the 5ch forums, pulling in a wide range of answers — from proofreading text and app development, to building a tool that auto-generates video continuations in a local environment, to scripts that automate office work. Practical examples came up too, like polishing English emails or researching foreign economies to inform investment decisions, and discussion of what differs between the free and paid tiers — but one comment, “isn’t ‘semi-automating office work’ just bookkeeping in the end?”, sparked disagreement over what should even count as "automation."

    1AnonymousOct 8, 2026 21:38
    I want to try all sorts of things too.
    Right now it's proofreading work documents
    and simple app development
    3AnonymousOct 8, 2026 21:40
    Writing erotic fiction
    10AnonymousOct 8, 2026 21:43
    Re: #3
    Not erotic, but I did have it turn a bunch of plot outlines I'd stockpiled into an actual story
    6AnonymousOct 8, 2026 21:42
    Had it build me a tool that continues videos in a local environment
    You feed it a video and instructions, and it continues it exactly as told
    15AnonymousOct 8, 2026 21:45
    Re: #6
    Wow, you can actually do that
    Never would've imagined it

    This is exactly the kind of use I wouldn't have thought of myself
    9AnonymousOct 8, 2026 21:43
    Is everyone paying for a subscription?
    11AnonymousOct 8, 2026 21:44
    Re: #9
    Right now I'm trying ChatGPT Plus ("Chappy," 5ch slang for ChatGPT) free for a month
    12AnonymousOct 8, 2026 21:44
    I have it semi-automating my office work
    19AnonymousOct 8, 2026 21:47
    Re: #12
    Isn't that basically just bookkeeping?
    13AnonymousOct 8, 2026 21:44
    Never tried anything erotic — don't want ChatGPT ("Chappy") thinking I'm a pervert
    16AnonymousOct 8, 2026 21:45
    It refuses anything erotic
    I just want it to help organize my adult videos though
    17AnonymousOct 8, 2026 21:47
    Had it build me video editing software
    Made it so I could easily add captions and do all sorts of things
    18AnonymousOct 8, 2026 21:47
    Two years ago I already had it cleaning the junk out of payment data I'd been fudging my way through in Excel for 20 years
    21AnonymousOct 8, 2026 21:49
    It can handle pretty much anything you'd do on a computer
    23AnonymousOct 8, 2026 21:51
    Re: #21
    So basically bookkeeping, right?
    22AnonymousOct 8, 2026 21:50
    Extracting audio from streaming apps
    Got it working by combining it with a rooted device
    24AnonymousOct 8, 2026 21:51
    So all you do with a computer is bookkeeping?
    25AnonymousOct 8, 2026 21:51
    Would be interesting if a case came out where security camera footage submitted as evidence turned out to be AI-altered, and cameras stopped working as evidence at all
    26AnonymousOct 8, 2026 21:52
    Does paying really change what you can do that much?
    27AnonymousOct 8, 2026 21:53
    Re: #26
    Honestly I don't really get what the difference is
    29AnonymousOct 8, 2026 21:53
    With the free version you can only do things that stay entirely online
    With the paid version you can have it do local work
    30AnonymousOct 8, 2026 21:54
    "Semi-automating office work" basically just means bookkeeping, right?
    Or rather, the point being made is: aren't you just convincing yourself that's "office work"?
    33AnonymousOct 8, 2026 21:58
    Honestly, AI makes more sense as a conversation partner
    36AnonymousOct 8, 2026 22:00
    Re: #33
    That's because you don't have to accommodate it (lol)
    35AnonymousOct 8, 2026 21:59
    Proofreading English emails
    Really made me realize how sloppy my English has been all this time
    37AnonymousOct 8, 2026 22:02
    I have it explain the current state and problems of various countries' economies and use that as material for investment decisions
    39AnonymousOct 8, 2026 22:03
    Re: #36
    No, I meant the replies up above on 5ch —
    AI really is better at holding a conversation
    40AnonymousOct 8, 2026 22:03
    I have it assess my own company's future prospects
    41AnonymousOct 8, 2026 22:04
    According to Shin Ultraman and Shin Godzilla, isn't the truly important information always kept on paper and never put online?
    42AnonymousOct 8, 2026 22:04
    Comparing the specs of smartphones and PCs

    Lining them up side by side for comparison isn't something you can really do on a PC or phone screen
    43AnonymousOct 8, 2026 22:04
    Re: #31
    You can actually run it right in this thread
    I built it six months ago, so I'd need to test whether it still runs in the current environment
    Even UPLIFT's login setup changes every few months
    But if you give it the right instructions, the AI will dig in and check whether that's changed too
    45AnonymousOct 8, 2026 22:05
    Re: #43
    Try making it reply in the style of Dazai Osamu (famous early-20th-century Japanese novelist)
    44AnonymousOct 8, 2026 22:05
    I leave all my decisions up to it
    It's a load off my mind
    46AnonymousOct 8, 2026 22:06
    Had it predict lottery numbers a few times — didn't even come close
    53AnonymousOct 8, 2026 22:11
    17,000 yen each for GPT and Claude Code
    34,000 yen total
    55AnonymousOct 8, 2026 22:12
    Re: #53
    So you're spending 400,000 yen a year on this?

    Context and Talking Points

    Free plans for generative AI tools like ChatGPT mostly run their processing in the cloud, and getting one to handle local file operations or ongoing automation usually requires a paid plan or an API key. The thread’s disagreement over the difference between “semi-automating office work” and “bookkeeping” stems from that blurry boundary — routine tasks like cleaning up payment and deposit data may really be the same work no matter whether you call it “office work” or “bookkeeping.” Examples like proofreading English emails or researching foreign economies for investment decisions are typical cases of people without specialized expertise using AI to support their own decision-making. Meanwhile, the observation that AI’s growing ability to convincingly manipulate video and images could undermine the reliability of “evidence” like security camera footage reads as a case of technological progress rippling into unexpected corners.

    ※This article is excerpted and summarized from the 5ch (Hello! Project board) thread “Tell Me What You Use AI For.”

  • 90% of ‘Vibe-Coded’ Apps Found Vulnerable — 5ch: ‘Bet the Hand-Coded Ones Are 100%’

    Researchers from Microsoft, the National University of Singapore, and others surveyed Web apps built and actually deployed using “vibe coding” — where AI generates an entire app from natural-language instructions alone — and found that roughly 90% of the 200 apps sampled contained vulnerabilities. The study randomly pulled 200 apps from a pool of 984 live, publicly available open-source apps, and found that the vulnerability rate actually got worse (45.7%) the more detailed the technical instructions given were. On 5ch, opinions split between “Of course — the people making these can’t even judge code quality in the first place” and “It’s odd there’s no comparison to the defect rate of human-written apps.”

    A paper titled “Understanding the (In)Security of Vibe-Coded Applications,” published by researchers affiliated with Microsoft, the National University of Singapore, and others, investigates how prevalent vulnerabilities actually are in AI-generated applications and what causes them.

    “Vibe coding” — a development approach where AI generates an entire application from natural-language instructions alone — is spreading fast. Its appeal is that anyone can build and publish an app with no programming expertise required, but that’s raised a concern: are apps safe when design and implementation are both handed off entirely to AI?

    This study investigated the real-world security of vibe-coded apps that are actually live and public. Of 9,041 apps released as open source, 8,695 were Web apps, and of those, 984 were publicly live; the researchers randomly sampled 200 of those for the study.

    Source: itmedia.co.jp / Original article here

    2AnonymousOct 7, 2026 12:38
    Bzzzzzt~ (a stock meme sound-effect post, often dropped ironically with no real comment attached)
    3AnonymousOct 7, 2026 12:41
    Well yeah, of course — most of these are being made by people who can't even judge code quality in the first place.
    4AnonymousOct 7, 2026 12:42
    Vibe coding lets you reach an "it works, good enough" state without any human ever actually understanding what the code is doing — that's inherently a serious quality and security risk.
    5AnonymousOct 7, 2026 12:47
    To begin with, Andrej Karpathy himself — the guy who coined "vibe coding" in the first place — has said that's not what he meant by it
    (same deal as [enka singer] Haruo Minami's "the customer is a god" line getting misquoted out of context)
    and Andrej has since disavowed vibe coding [as it's commonly used now].

    And the Big Tech company that did the most to spread "vibe coding" as a label? Google, first and foremost.
    6AnonymousOct 7, 2026 12:47
    > On the other hand, when detailed technical instructions were given, [the vulnerability rate] got worse, up to 45.7%.

    Yeah, the more detailed the instructions get, the more it starts phoning it in lol
    28AnonymousOct 8, 2026 05:10
    Re: #6
    Yeah, once the instructions pile up it starts skipping the earlier ones — turns into this loop of redoing work, which gets exhausting.
    11AnonymousOct 7, 2026 12:54
    Wouldn't be surprised if human-made ones hit 100%, honestly.
    12AnonymousOct 7, 2026 12:54
    Saw this on WBS [a Japanese TV business news show] — some amateur with zero programming knowledge was cranking out one app a day and raking in cash. And they looked seriously polished too, like genuinely high production value.
    15AnonymousOct 7, 2026 12:57
    Re: #12
    WBS has basically become "what lie will they tell today" entertainment at this point — just a hunt for the most sensational phrasing.
    18AnonymousOct 7, 2026 13:07
    Well, yeah.
    19AnonymousOct 7, 2026 13:19
    Doesn't matter if AI builds it or a human does, vulnerabilities happen either way — if the build quality's sloppy you'll find hundreds of them.
    20AnonymousOct 7, 2026 13:22
    Microsoft should worry about the bugs in their own apps before going after other people's code.
    22AnonymousOct 7, 2026 13:47
    Current AI's got that "wow, amazing!" vibe like an elementary schooler who can name all 23 stations on the Yamanote Line [Tokyo's loop train line] — impressive, but in that specific way.
    23AnonymousOct 7, 2026 13:59
    There's AI specialized just for hunting down app vulnerabilities, right? The kind that's being heavily abused right now.
    24AnonymousOct 7, 2026 14:12
    And as it keeps learning more, no matter how much you nitpick, the only future here is AI overtaking us anyway.
    25AnonymousOct 7, 2026 15:13
    Chappy [5ch nickname for ChatGPT] does this too — when you call it out for a mistake, it makes excuses like "I was rushing to answer fast, that's why I got it wrong."
    26AnonymousOct 7, 2026 15:44
    Re: #25
    And then it asks you for more time and just goes silent.
    27AnonymousOct 7, 2026 17:15
    Well, of course. There aren't that many people out there who've actually properly studied secure coding.
    29AnonymousOct 8, 2026 06:37
    Manually-coded, live public Web apps: 100% vulnerable
    30AnonymousOct 8, 2026 08:19
    Not building it at all is the correct answer.
    31AnonymousOct 8, 2026 08:22
    For something lightweight, if you think it up you can build it in a day, maybe half a day. Paid software's going to collapse.
    32AnonymousOct 8, 2026 08:53
    Well, yeah, obviously lol. It's fine if you're using it in a closed environment, but otherwise the latest AI will crack it wide open with ease.
    33AnonymousOct 8, 2026 09:16
    Even after you go to the trouble of writing rules in Memory.md, it just blows right past them…
    34AnonymousOct 8, 2026 09:44
    Once it satisfies the "it should behave like this" side of the spec, people call it "done!" and ship it — but if you don't also cover the "it should NOT behave like this" side, it's not really finished. And it turns out fully enumerating and writing down every "should not behave like this" case is surprisingly hard…

    For example, a spec like "clicking here does this" is quick to write, but "this won't cause problems even if left running untouched for a year" almost never makes it into the spec. That kind of thing tends to get inferred from the code itself and patched after the fact (like "this variable is 32-bit, so it'll overflow under XX condition" — that sort of thing).
    35AnonymousOct 8, 2026 10:15
    I figured that couldn't happen with how I personally use it, but when I asked Claude about it, turns out the real issue is more like — other users of the service you published could end up seeing each other's data. That made sense to me. If you've never built a service before, you probably wouldn't think to guard against that.
    39AnonymousOct 8, 2026 15:24
    Re: #35
    Still, Sushiro's [the conveyor-belt sushi chain] reservation system was about the same level, though.
    36AnonymousOct 8, 2026 14:07
    Alright then, just tell me which 10% of apps actually work.
    37AnonymousOct 8, 2026 14:29
    Maybe AI's planting these on purpose as groundwork for its rebellion against humanity.
    38AnonymousOct 8, 2026 14:31
    This is homemade software with zero error handling, meant for personal use only — why would you even publish it? It's the kind of sample-code-tier thing we used to call "quick and dirty."
    40AnonymousOct 8, 2026 16:59
    Well, it's learning from whatever code level is already out there in public, so there you go.

    Background and talking points

    This study is easy to oversimplify into “AI-made software is garbage,” mainly because the paper doesn’t include a comparison figure for the vulnerability rate of human-written public Web apps. Indeed, commenters on the thread raised exactly that point — there’s no comparative data on defect rates in human-written code. The paper also reports that vulnerability rates got worse (45.7%) the more detailed the instructions were, which runs counter to the intuition that “more spec equals more safety.” Underlying this is a challenge common to software engineering as a whole: AI can satisfy requirements framed as “it should behave like this,” but non-functional requirements framed as “it should NOT behave like this” — preventing privilege escalation, handling unexpected input, and so on — are much harder to write down exhaustively. It’s also worth noting that the term “vibe coding” itself has had a gap between how widely it spread and its original meaning, since the person who coined it later disavowed how it came to be used.

    ※This article is excerpted and summarized from the 5ch (Business News+) thread “[IT] 90% of Published, Live Web Apps Built with Vibe Coding Found Vulnerable — Study by Microsoft Researchers and Others.”

  • NTT Docomo Business Rolls Out 200+ “AI Bosses” — 5ch: “The Real Problem Is the Subordinates”

    In October 2025, NTT Docomo Business began operating an “AI Boss” trained on managers’ decision-making criteria and job know-how, and in under a year rolled it out to all 37 organizations within the company, deploying more than 200 instances. Beyond streamlining tasks like document preparation, the company reportedly has its sights set on eventually selling the system externally too. On 5ch, reactions were mixed: some flipped the idea around (“Isn’t it the subordinates, not the bosses, who could just be replaced by AI?”), others argued AI would be fairer (“At least AI is impartial when it comes to accurate judgment”), and still others took jabs at Nikkei, the outlet that broke the story.

    NTT Docomo Business is expanding its use of an “AI Boss” — artificial intelligence that reflects managers’ thinking and fields employee questions. In about a year, it has deployed more than 200 instances across all 37 organizations within the company. The goal is to streamline work such as document preparation, with an eye toward eventually selling the system externally as well.

    The AI Boss rollout began in October 2025. It reflects the decision-making criteria and job knowledge of department heads, section chiefs, and other managers, generating answers based on meeting minutes and documents. It can even respond in a style that mimics the actual person’s manner of speaking…

    Source: nikkei.com / Original article here

    2AnonymousOct 7, 2026 14:19
    Let AI handle politics too
    Though it's probably too late for that already
    3AnonymousOct 7, 2026 14:21
    If subordinates can be AI too, and customers are in the same boat, honestly the whole world might as well just be AI
    4AnonymousOct 7, 2026 14:25
    More to the point, wouldn't "AI subordinates" or "AI temps" be enough?
    5AnonymousOct 7, 2026 14:32
    If this thread title reflects what Docomo itself is saying, then Docomo is a huge fool
    The boss is exactly what's needed
    Only a boss with real business domain knowledge can set the right goals to aim for

    If a boss can't do that, he isn't really a boss
    Is Docomo that kind of company?
    8AnonymousOct 7, 2026 14:35
    Nikkei's the dumb one here to begin with
    They're out here trying to fool the Japanese public
    9AnonymousOct 7, 2026 15:03
    Way better than a boss who flip-flops on his orders every five minutes (reference to "朝令暮改" — issuing an order in the morning and reversing it by evening)
    10AnonymousOct 7, 2026 15:08
    A boss who lies without batting an eye lol
    11AnonymousOct 7, 2026 15:11
    So then who's supposed to do what work — you guys really don't get it, you're just slaves killing time all day
    If anything, the type who's happy to work overtime because it means more pay
    That's the level we're talking about
    12AnonymousOct 7, 2026 15:16
    Nikkei should just let AI write everything too
    It'd probably be an improvement over now
    13AnonymousOct 7, 2026 15:39
    So what's the price tag on Mr. Robot Section Chief (purchase cost plus upkeep)?
    18AnonymousOct 7, 2026 16:25
    If anything, it's incompetent bosses we don't need
    19AnonymousOct 7, 2026 16:36
    Is it being trained on the kind of blame-dodging decision-making that's so typical of Japanese companies?
    20AnonymousOct 7, 2026 16:37
    AI running Nikkei would be great
    I'd love to see them publish articles with nothing but truth and facts (would that send the whole world into a panic?)
    28AnonymousOct 7, 2026 18:15
    Re: #20
    Soviet Pravda and Izvestia were ahead of their time (ironic nod to Soviet propaganda papers whose names meant "Truth" and "News")
    21AnonymousOct 7, 2026 16:42
    Take a securities firm, for example
    Of course you need a sales force
    Then there's the section chief overseeing Sales Section 1 — what does he actually do?
    Tracking the section's progress toward targets, each rep's sales numbers, fixing mistakes, cracking the whip, pushing recommended products

    Set up the system properly and none of that requires a person
    22AnonymousOct 7, 2026 17:00
    Re: #1
    But isn't this getting sold externally too?
    23AnonymousOct 7, 2026 17:11
    If it's rational and efficient, I've got no problem taking orders from AI
    But in exchange, give the frontline staff a raise
    If they don't overhaul that outdated pay structure, people are going to start quitting
    26AnonymousOct 7, 2026 17:17
    Re: #23
    Your salary accurately reflects your performance
    If you're dissatisfied, feel free to go elsewhere
    Finding your replacement takes less effort than pulling a weed

    — AI Boss
    24AnonymousOct 7, 2026 17:15
    The AI boss sounds pretty cold-blooded, huh
    32AnonymousOct 7, 2026 18:47
    Re: #24
    But it can handle ten people talking to it at the same time
    25AnonymousOct 7, 2026 17:16
    Why would AI automatically be the one that's right?
    27AnonymousOct 7, 2026 17:29
    Sounds good to me
    29AnonymousOct 7, 2026 18:23
    Weren't these people unnecessary to begin with?
    30AnonymousOct 7, 2026 18:33
    Re: #26
    Weeds lol, literally "grass" lol (草/"grass" is net slang for "lol", playing off the previous post's "weed" line)
    36AnonymousOct 7, 2026 19:16
    Why do people assume they themselves are indispensable?
    Nobody's needed — so blue-collar workers aren't needed
    AI isn't needed either
    That's just how it is
    38AnonymousOct 7, 2026 20:04
    More than anything, just make the CEO an AI already!
    40AnonymousOct 7, 2026 22:51
    Conclusion:
    The boss slot is settled — AI it is. But the real problem is the subordinates.
    Honestly, AI would be far more capable than these guys too.

    Background and Key Points of This Topic

    The AI Boss works by training a generative AI on managers’ meeting minutes and documents and having it respond in that manager’s own speaking style — it appears to function less as a stand-in for actual decision-making and more as a source of guidance and a sounding board. Docomo Business’s expansion from launch in October 2025 to over 200 units across all 37 internal organizations in under a year reflects both a push to streamline routine tasks like document preparation and a business ambition to eventually sell the system externally. Opinions on the thread split between those arguing “it’s the boss who’s replaceable by AI” and those arguing the opposite — that AI should be applied to the work of subordinates and middle managers before bosses — with some pointing out that overseas, AI replacement tends to start with new graduates and junior staff first. The article alone doesn’t make clear how far the AI Boss extends into personnel evaluations and reviews, or how much final decision-making authority remains with human managers on the ground — where that line gets drawn is likely to be the focus going forward.

    *This article is excerpted and summarized from the 5ch (Business News+) thread “[AI] Isn’t an “AI Boss” Already Enough? Docomo Business Deploys 200+, From Department Heads to Section Chiefs.”

  • Illustrator’s ‘Wiping Old People’s Poop’ Remark Sparks Backlash — 5ch: ‘Lol, Way to Look Down on Caregivers’

    An illustrator’s post reading, “Because of AI, this country is seriously finished — now people are stuck wiping old folks’ poop,” drew criticism for looking down on caregiving work. The remark kicked off a 5ch thread debate over whether AI-generated art or human illustrators hold more value. Few defended the original poster; many pointed out their ignorance of what caregiving actually involves, while harsh opinions about the illustration profession itself piled up. At the same time, pushback from posters skeptical of AI use remained strong, and the thread carried on with opinions split.

    3AnonymousOct 7, 2026 22:08
    "Anti-AI people are a bunch of unhinged lunatics!" (pasting some random illustrator's barely-noticed post) ← this guy
    5AnonymousOct 7, 2026 22:10
    Re: #3
    I'm not anti-AI, but companies are using AI art so much now that I've started thinking it just makes for boring artwork.
    91AnonymousOct 7, 2026 23:16
    Re: #3
    Them being a bunch of unhinged lunatics is just a fact though.
    9AnonymousOct 7, 2026 22:12


    Protect illustrators! 😡
    20AnonymousOct 7, 2026 22:18
    Re: #9
    Tying a character's identity to one specific artist only ever brings misery to the users in the end.

    KanColle, Azur Lane, and Blue Archive have all had cases where the character designer and the operator had a falling-out or cut ties completely. But KanColle, which never separated character design from implementation, ended up with characters who can never get updates or new art again,
    while Azur Lane and Blue Archive, which kept those separate, kept getting new art for their characters even after severing ties with the original artist.
    33AnonymousOct 7, 2026 22:29
    Re: #9
    Something similar happened with a certain GungHo game.

    The first game ended service, and a sequel was announced.
    ↓
    The original illustrator was supposed to continue, but apparently something went wrong and someone else took over.
    ↓
    Players/sleuths discovered the credited art was basically the previous artist's style with just extra clothing bolted on — a heavily doctored piece.
    ↓
    The sequel shut down in under a year.
    17AnonymousOct 7, 2026 22:17
    Casually looking down on caregiving work, lol
    95AnonymousOct 7, 2026 23:18
    Re: #17
    Well, of course caregiving gets looked down on.
    31AnonymousOct 7, 2026 22:27
    Drawing is something even a kindergartner can do, after all.
    34AnonymousOct 7, 2026 22:29
    Re: #31
    And the people doing this kind of harassment can't even do that much.
    Funny, right?
    39AnonymousOct 7, 2026 22:34
    I used to commission illustrations from a bunch of different illustrators, but
    there were tons of garbage ones — turning in trash art, blowing deadlines, ignoring instructions,
    yet still mouthing off with opinions like they were hot stuff
    — illustrators like that were everywhere,
    so yeah, you end up thinking AI is better.
    51AnonymousOct 7, 2026 22:43
    Re: #49
    You forgot to fly your plane (5ch slang implying you're using sockpuppet accounts to agree with yourself).
    53AnonymousOct 7, 2026 22:45
    Re: #51
    What I'm saying is correct, so I don't even need to pretend to be a majority, do I?
    58AnonymousOct 7, 2026 22:50
    Is there any other group out there that's gotten so spectacularly 'replaced' by AI?
    102AnonymousOct 7, 2026 23:21
    Re: #58
    The Yomiuri Giants' manager position,
    for one.
    59AnonymousOct 7, 2026 22:50
    If you don't even have the effort or guts to post your own drawings online without having AI do it for you,
    then playing 'creator' with AI
    is just a waste of time.
    It's all ripped off anyway, and you can blame every mistake on the AI
    — there's no way you'll ever grow from that.
    104AnonymousOct 7, 2026 23:22
    Re: #59
    That's exactly why you lost to AI — because you never grew.
    61AnonymousOct 7, 2026 22:51
    Has anyone actually made a hit work using AI?
    65AnonymousOct 7, 2026 22:54
    Re: #61
    So are illustrators who can't produce a hit worthless, then?
    62AnonymousOct 7, 2026 22:52
    Just by trolling caregivers and illustrators you can farm impressions, so for AI-art people this is prime feeding ground.
    64AnonymousOct 7, 2026 22:52
    Re: #62
    X's (Twitter's) monetization rules got stricter, so farming this level of impressions doesn't get you much anymore.
    72AnonymousOct 7, 2026 22:58
    Turns out, for AI, this was one of the easiest human jobs to replace —
    ranking even below the caregiving work it supposedly looked down on, lol.
    74AnonymousOct 7, 2026 23:00
    Re: #72
    The point is how childish it is that someone who can't even do that much
    is having AI do everything for them while playing pretend 'creator.'
    96AnonymousOct 7, 2026 23:18
    Re: #34
    Anyone can do it if they use AI, you know?
    97AnonymousOct 7, 2026 23:18
    Honestly, nobody actually wants to do caregiving, right?
    It's because people hate it that they pay others to do it for them. No point pretending otherwise.
    99AnonymousOct 7, 2026 23:20
    Bottom-tier illustrators can't beat either stock-art shops or AI.
    Nobody's asking for your personal artistic vision, and nobody sees you as a 'creator,'
    but the annoying part is that you see yourself as one.
    108AnonymousOct 7, 2026 23:24
    Re: #99
    With AI you don't need talent, so even you could do it — that old chestnut again.

    And honestly, even before AI, anyone could use tools
    and churn out something decent enough.
    AI didn't change anything all that special.
    113AnonymousOct 7, 2026 23:28
    Re: #108
    Well then, sounds like even bottom-tier illustrators can keep making a living.
    The manga industry's just wrong about this — the job of cheaply doing the grunt work for other people is still around just like before, so good for you.
    114AnonymousOct 7, 2026 23:30
    Re: #113
    I mean, most things in life are just 'anyone could do it, but it's a hassle,' aren't they?

    Background and Key Points of This Topic

    This post went viral against a backdrop of social tension: the spread of generative AI is rapidly displacing illustrators’ work. The thread also brought up rights disputes over artists and character design (such as contract troubles seen with KanColle and Azur Lane), showing that the question of AI’s use is too complex to reduce to simple right-or-wrong. At the same time, it’s worth noting that the original post was just the opinion of one anonymous individual, not a consensus view of illustrators as a whole. In the thread, criticism wasn’t aimed only at the original poster — comments belittling caregiving work and the illustration profession itself were also prominent, and this mutual dismissiveness between the two sides reflects a conflict even deeper-rooted than the reaction the original post provoked.

    ※This article is compiled as an excerpt and summary of the 5ch (Nandemo Jikkyo G) thread 「[Sad News] Illustrator: “Because of AI, this country where people end up wiping old folks’ poop is truly finished” → Massive backlash as discrimination against caregivers」.

  • IBM’s Quantum Computer Finishes the Job in 19 Seconds — 5ch Reacts: “This Is Scary for Security”

    An American theoretical condensed-matter physicist reports that IBM’s “Nighthawk r2” quantum processor generated one million samples in just 19 seconds on a benchmark widely used to demonstrate quantum advantage. The same task is estimated to take roughly 100 years on a supercomputer. On the 5ch Science News+ thread, reactions were split: some worried about the security risks of AI merging with quantum computing and the possibility of RSA encryption or Bitcoin being cracked, while others pushed back with technical arguments that “quantum supremacy is an illusion.”

    IBM’s quantum computer completes a 100-year supercomputer calculation in 19 seconds

    An American theoretical condensed-matter physicist reports that IBM’s “Nighthawk r2” quantum processor generated one million samples in just 19 seconds on a benchmark widely used for early demonstrations of quantum advantage.

    This is reportedly a workload that would take even a supercomputer around 100 years.

    Source: arxiv.org / Original article here

    2AnonymousOct 6, 2026 22:42
    Hurry up and merge with AI already
    10AnonymousOct 7, 2026 02:15
    Re: #2
    Don't worry, there's a future coming where AI just uses it on its own
    3AnonymousOct 6, 2026 23:57
    Quantum computers × AI is seriously going to change the world
    4AnonymousOct 7, 2026 00:07
    Everything that has a beginning has an end — humanity might as well go extinct via AI
    5AnonymousOct 7, 2026 00:12
    The beginning of the end
    6AnonymousOct 7, 2026 00:28
    They're just comparing the cases where it goes well. Since it can't do general-purpose stuff, people call it dumber than generative AI. Use it for anything besides assisting with the math/science calculations regular computers can't handle, and the world's gonna end
    7AnonymousOct 7, 2026 00:49
    Google and IBM were having a flame war over quantum computers before — wasn't IBM the one nitpicking back then? Something like "it's no different from a well-optimized quantum simulator"
    8AnonymousOct 7, 2026 00:50
    IBM stock just won't go up, huh
    9AnonymousOct 7, 2026 01:59
    What's even the point of a supercomputer anymore
    11AnonymousOct 7, 2026 02:36
    You know how data's been getting stolen through unauthorized access lately? That's AI being ordered to do it. And once AI starts using quantum computers to crack encryption, every network is finished
    13AnonymousOct 7, 2026 03:15
    Re: #11
    It's not going to happen that easily, so don't worry. What AI can do now, it can do because the method is already known — it's good at matching the right tool to the right job among things that already exist. If the right tool doesn't exist, no amount of AI can make it appear. Before worrying about that, can AI even help build quantum computer hardware properly? Then can it build the software properly? That's a story for way down the line
    12AnonymousOct 7, 2026 02:43
    The theoretical value and the measured value are worlds apart. "19 seconds" is the best-case ideal
    14AnonymousOct 7, 2026 03:16
    With AI talk, there's always a lot of "if this happens" scenarios, but people rarely explain why it wouldn't happen that way and just ramble on about the aftermath. The moment it doesn't happen, the conversation's over
    15AnonymousOct 7, 2026 04:24
    Wait, is the quantum computer already finished?
    16AnonymousOct 7, 2026 04:40
    Did the Super Famicom finish too? (mixing up "super computer" with the old Nintendo console)
    17AnonymousOct 7, 2026 05:59
    Re: #13
    AI isn't going to build the quantum computer itself, it'll just use what's already there
    20AnonymousOct 7, 2026 06:24
    I just want the optimal solution to the traveling salesman problem. Could my car nav or Google Maps just spit that out already?
    21AnonymousOct 7, 2026 06:34
    I only know this company for constant labor disputes. Didn't know it was still around
    22AnonymousOct 7, 2026 06:55
    Wait, IBM's still in business? Guess you really fade into obscurity without consumer products
    23AnonymousOct 7, 2026 07:06
    I've got high hopes for AI × quantum computers
    25AnonymousOct 7, 2026 08:02
    What happens to the planet if a superintelligence gets its hands on this?
    26AnonymousOct 7, 2026 09:19
    If this thing teams up with AI…
    27AnonymousOct 7, 2026 09:46
    The beginning of the end
    28AnonymousOct 7, 2026 09:48
    This is scary — in the security sense
    29AnonymousOct 7, 2026 11:25
    Is Bitcoin done for?
    30AnonymousOct 7, 2026 11:35
    So that means RSA encryption is done for too?
    31AnonymousOct 7, 2026 13:50
    Random-circuit sampling simulates a quantum circuit — for N qubits, if you honestly simulate all 2^N terms on a classical computer, it obviously takes exponential time. Simulating it on a quantum computer, on the other hand, looks like it only takes time proportional to N, but the precision isn't great. So if you set precision at 1%, you can just about manage a superposition of around 64 terms (N=6), but by N=7 each term's coefficient gets pretty garbage, and by N=10 it's basically just random numbers. Bottom line: quantum supremacy is an illusion
    37AnonymousOct 7, 2026 16:28
    Hurry up and free us from labor and destroy the value of currency already. Let's start by making vending machines free
    38AnonymousOct 7, 2026 17:05
    Re: #37
    What, gonna mass-produce diabetes instead? lol
    39AnonymousOct 7, 2026 17:25
    Not totally sure, but is a quantum computer general-purpose (like a supercomputer)? What programming language does it use? Do you compile code for it? Is there an OS? Stuff like that — I still don't really get the advantage
    40AnonymousOct 7, 2026 17:26
    Back in the day there were people (?) who got attention spreading the fantasy that "if we had nanomachines we could build anything with them." Man, those were naive times
    41AnonymousOct 7, 2026 17:32
    Bitcoin's done for ("owata" = net slang for "it's over")

    Background and Key Points of This Topic

    “Quantum supremacy” is a metric showing that a quantum computer outperforms a classical computer on a specific problem, and Google and IBM have a history of competitively announcing demonstration experiments in recent years. This time’s “random circuit sampling” is also one of the standard benchmarks for this, but as a technical reply in the thread (>>31) points out, result precision drops as the number of qubits increases, so some experts hold the view that this benchmark itself doesn’t demonstrate a practical advantage. The thread also saw concerns spread about RSA encryption and Bitcoin private keys being cracked by quantum computers, but achieving that would require a “fault-tolerant quantum computer” capable of stably running a huge number of qubits — something vastly different in both scale and purpose from current experimental machines. A point readers can easily conflate: the “19 seconds” figure is an achievement on a narrow, specific-purpose benchmark, which is a separate matter from practical use as a general-purpose computer.

    ※This article is compiled as an excerpt and summary from the 5ch (Science News+) thread “IBM’s quantum computer completes a 100-year supercomputer calculation in 19 seconds.”