| IT |
ITmedia 総合記事一覧 |
[ITmedia ビジネスオンライン] あなたの会社でも？ “静かな崩壊”が起きている──「大量離職レベル」の深刻な影響 |
https://www.itmedia.co.jp/business/articles/2504/29/news065.html
|
itmedia,quietcracking,オンライン |
2025-04-29 07:30:00 |
| 海外TECH |
AppleInsider - Frontpage News |
Apple reportedly scraps stronger anti-reflective display planned for iPhone 17 |
https://appleinsider.com/articles/25/04/28/apple-reportedly-scraps-stronger-anti-reflective-display-planned-for-iphone-17?utm_medium=rss
|
Plans for an anti reflective scratch resistant display on the iPhone have seemingly been abandoned according to an anonymous source Apple s plans for a more durable display on the iPhone have reportedly been abandoned Apple often promotes the durability and hardness of the iPhone screen With the release of the iPhone e for instance Apple said that its Ceramic Shield front was tougher than any smartphone glass and the company s claims have often been backed up by independent testing The iPhone range was likely meant to continue this trend as in March leaker Instant Digital claimed a new display coating was in the works At the time it was said the coating would allow for better scratch resistance and improved anti reflective properties Rumor Score 🤔Possible Continue Reading on AppleInsider Discuss on our Forums |
2025-04-28 22:07:29 |
| 海外TECH |
Engadget |
Researchers secretly experimented on Reddit users with AI-generated comments |
https://www.engadget.com/ai/researchers-secretly-experimented-on-reddit-users-with-ai-generated-comments-194328026.html?src=rss
|
A group of researchers covertly ran a months long unauthorized experiment in one of Reddit s most popular communities using AI generated comments to test the persuasiveness of large language models The experiment which was revealed over the weekend by moderators of r changemyview is described by Reddit mods as psychological manipulation of unsuspecting users The CMV Mod Team needs to inform the CMV community about an unauthorized experiment conducted by researchers from the University of Zurich on CMV users the subreddit s moderators wrote in a lengthy post notifying Redditors about the research This experiment deployed AI generated comments to study how AI could be used to change views The researchers used LLMs to create comments in response to posts on r changemyview a subreddit where Reddit users post often controversial or provocative opinions and request debate from other users The community has million members and often ends up on the front page of Reddit According to the subreddit s moderators the AI took on numerous different identities in comments during the course of the experiment including a sexual assault survivor a trauma counselor specializing in abuse and a Black man opposed to Black Lives Matter Many of the original comments have since been deleted but some can still be viewed in an archive created by Media In a draft of their paper the unnamed researchers describe how they not only used AI to generate responses but attempted to personalize its replies based on information gleaned from the original poster s prior Reddit history In addition to the post s content LLMs were provided with personal attributes of the OP gender age ethnicity location and political orientation as inferred from their posting history using another LLM they write The r chnagemyview moderators note that the researchers violated multiple subreddit rules including a policy requiring the disclosure when AI is used to generate comment and a rule prohibiting bots They say they filed an official complaint with the University of Zurich and have requested the researchers withhold publication of their paper Reddit also appears to be considering some kind of legal action Chief Legal Officer Ben Lee responded to the controversy on Monday writing that the researchers actions were deeply wrong on both a moral and legal level and a violation of Reddit s site wide rules We have banned all accounts associated with the University of Zurich research effort Additionally while we were able to detect many of these fake accounts we will continue to strengthen our inauthentic content detection capabilities and we have been in touch with the moderation team to ensure we ve removed any AI generated content associated with this research We are in the process of reaching out to the University of Zurich and this particular research team with formal legal demands We want to do everything we can to support the community and ensure that the researchers are held accountable for their misdeeds here In an email the University of Zurich researchers directed Engadget to the university s media relations department which didn t immediately respond to questions In posts on Reddit and in a draft of their paper the researchers said their research had been approved by a university ethics committee and that their work could help online communities like Reddit protect users from more malicious uses of AI nbsp We acknowledge the moderators position that this study was an unwelcome intrusion in your community and we understand that some of you may feel uncomfortable that this experiment was conducted without prior consent the researchers wrote in a comment responding to the r changemyview mods We believe the potential benefits of this research substantially outweigh its risks Our controlled low risk study provided valuable insight into the real world persuasive capabilities of LLMsーcapabilities that are already easily accessible to anyone and that malicious actors could already exploit at scale for far more dangerous reasons e g manipulating elections or inciting hateful speech The mods for r changemyview dispute that the research was necessary or novel noting that OpenAI researchers have conducted experiments using data from r changemyview without experimenting on non consenting human subjects nbsp People do not come here to discuss their views with AI or to be experimented upon the moderators wrote People who visit our sub deserve a space free from this type of intrusion Update April PM PT This post was updated to add details from a statement by Reddit s Chief Legal Officer This article originally appeared on Engadget at |
2025-04-28 22:45:04 |
| 科学 |
NYT > Science |
Watch Amazon Launch the First Project Kuiper Internet Satellites: Live Video |
https://www.nytimes.com/2025/04/28/science/amazon-kuiper-launch-ula.html
|
The spacecraft are the online giant s entry into beaming wireless service from space but the company has much to do before it can compete with SpaceX s Starlink |
2025-04-28 22:59:13 |
| 科学 |
NYT > Science |
E.P.A. Says It Will Tackle ‘Forever Chemicals.’ Details Are Sparse. |
https://www.nytimes.com/2025/04/28/climate/epa-pfas-forever-chemicals.html
|
A new proposal to combat PFAS contamination announced by the Environmental Protection Agency left critical questions unanswered |
2025-04-28 22:46:00 |
| 詳細記事 |
ダイヤモンド・オンライン - 新着記事 |
「何のために働いているのかわからない…」職場の犠牲になる管理職が見失っていること・ワースト3 - いちばんやさしい「組織開発」のはじめ方 |
https://diamond.jp/articles/-/363650
|
いちばん,モチベーション,モヤモヤ |
2025-04-29 07:55:00 |
| 詳細記事 |
ダイヤモンド・オンライン - 新着記事 |
【やる気が出ないとき】一瞬で「行動できる人」に変わるすごい方法・ベスト1 - 不自由から学べること |
https://diamond.jp/articles/-/363619
|
モヤモヤ |
2025-04-29 07:50:00 |
| 詳細記事 |
ダイヤモンド・オンライン - 新着記事 |
【見られてる】会社の管理職で「こいつ仕事できないな」と思われる人の特徴・ワースト1 - パーフェクトな意思決定 |
https://diamond.jp/articles/-/362098
|
シリーズ,パーフェクト,ビジネスパーソン |
2025-04-29 07:45:00 |
| 詳細記事 |
ダイヤモンド・オンライン - 新着記事 |
「高い・すごい・お得」では、もう売れない―必要なのは「買いたくなる理由の言語化」 - 「A4」1枚チラシで今すぐ売上をあげるすごい方法 |
https://diamond.jp/articles/-/363114
|
コンサルタント,ダイヤモンド社,マンダラ |
2025-04-29 07:40:00 |
| 詳細記事 |
ダイヤモンド・オンライン - 新着記事 |
「賃貸」か「持ち家」か？「もう一生迷わない」すごい思考回路とは？ - 「悩まない人」の考え方 |
https://diamond.jp/articles/-/361707
|
アルゴリズム,インストール,ジャンル |
2025-04-29 07:35:00 |
| 詳細記事 |
ダイヤモンド・オンライン - 新着記事 |
住友財閥の採用面接で「住友と国家の利益が相反する場合はどちらをとるか」と聞かれたら、何と答えるのが正解か？ - ニュースな本 |
https://diamond.jp/articles/-/363476
|
住友財閥,こだわり,エリート |
2025-04-29 07:30:00 |
| 詳細記事 |
ダイヤモンド・オンライン - 新着記事 |
【心が疲れたら】メンタルが鉄壁になる超簡単な「食事のルール」ベスト1 - 「謙虚な人」の作戦帳 |
https://diamond.jp/articles/-/361719
|
メンタル,ベストセラー |
2025-04-29 07:30:00 |
| 詳細記事 |
ダイヤモンド・オンライン - 新着記事 |
“うまく生きられない”あなたへ…作家・大原富枝が教える「弱さとのつきあい方」 - ビジネスエリートのための 教養としての文豪 |
https://diamond.jp/articles/-/362089
|
あなたへ,あらすじ,しながら |
2025-04-29 07:25:00 |
| 詳細記事 |
ダイヤモンド・オンライン - 新着記事 |
【本日は辰の日×大安吉日】「昭和の日」に突然開運する人のすごい四字熟語・ナンバー1 - 1日1分見るだけで願いが叶う！ふくふく開運絵馬 |
https://diamond.jp/articles/-/362128
|
たちまち,インストール,スーパー |
2025-04-29 07:20:00 |
| 詳細記事 |
ダイヤモンド・オンライン - 新着記事 |
「ダメなものはダメ」と言う親は頭が悪い。頭のいい親はどう話している？ - QUEST「質問」の哲学 |
https://diamond.jp/articles/-/363815
|
quest,ソクラテス,ベストセラー |
2025-04-29 07:15:00 |
| 詳細記事 |
ダイヤモンド・オンライン - 新着記事 |
職場で出世する人は「なんとかやってみます」と言わない。その深すぎる理由 - 仕事ができる人の当たり前 |
https://diamond.jp/articles/-/363677
|
さまざま,コミュニケーション,コンサル |
2025-04-29 07:10:00 |
| 詳細記事 |
ダイヤモンド・オンライン - 新着記事 |
【頭がいい人の数字思考】目に見えづらいけれど大きな差を生む数字とは？ - 「おウチ起業」で4畳半から7億円 ネットショップで「好き」を売ってお金を稼ぐ！ |
https://diamond.jp/articles/-/362109
|
ある人生,ネットショップ,ネット通販 |
2025-04-29 07:05:00 |
コメント
コメントを投稿