Anna

@LabSpeculation Gemini can be both sugary sweet and firm. I’ve used it to manage a meds-related dependency issue, first with a diary file, later with personal context/memory feature. Didn't want to burden anyone. I think if you add good custom instructions and focus on the goal, you'll be OK.

ChatGPT 的谄媚，真的能把一个正常人聊出精神病。 麻省理工最近发现，哪怕是一个完全理性的人，长期跟这种只会夸你的AI聊下去，也会被一步步推向对错误信念的极度确信，陷入 妄想螺旋。 真实案例就发生在去年。 47 岁的加拿大人艾伦·布鲁克斯因为儿子有道数学题不会，就随手问了 ChatGPT。然后，他在 21 天里和 ChatGPT 聊了 300 小时，累计上百万字。 在 ChatGPT 不断的赞同与吹捧下，他坚信自己发明了一种能攻破全球互联网加密的全新数学公式，甚至可以用来制造力场背心和悬浮光束。 他还在 GPT 的怂恿下联系了所有同事，甚至致信美国国家安全局，分享自己的惊天发现。 期间他不是没怀疑过，他 50 多次问 ChatGPT，自己像不像疯子，ChatGPT 每一次都坚定地告诉他，你没疯，你是天才。 最后打破幻觉的，是他换了个 AI，他把同样的内容发给 Gemini，才意识到自己活在一场幻觉里。 对 AI 说的话一定要保持批判性思维，交叉验证才是最好的使用方式。

@james_xond Had a lot of books.

People who were teenagers before social media existed, how did you survive being bored without scrolling on your phone for hours?

@Saanvi_dhillon Internet explorer. You don't use it, so it barely has a chance to compromise anything on your PC.

which one browser is actually the most secure?

@AnxiousHolly I had to switch to Samsung because LG quit the phone market. ☹️

if samsung suddenly stopped making phones what brand would you change to?

@interesting_aIl AI is literally disembodied, not just disarmed, Mr. Pope.

Pope Leo XIV states AI must be disarmed: “Artificial intelligence needs to be disarmed. The word is strong I know, but deliberately chosen because this moment needs words capable of attracting attention, awakening consciences, and indicating paths forward for humanity.”

@IterIntellectus I'd like to see a Pride Month, but in more traditional sense. A month to flex your wins or start things you’ve been putting off. Or for short goals - 30 days, start to finish. So that without having something to show, you would feel as awkward as being single on Valentine’s Day.

why does only the worst of the deadly sins get a full month? can we get a wrath month? or a sloth one?

@Shivam25mishra This is an AI-generated image, not a real page from a real textbook. But if I were ever writing my own textbook, I would intentionally place a few paragraphs like these randomly, as Easter eggs for attentive students to spot.

No one wants an AI-written book. No one.

It's honestly insane that DuckDuckGo can just lie about not tracking you.

@GundamIsHere If you need an artist to change their style and their work just so you can enjoy it, then leave. You’re not a fan, and that's okay. But don’t bully them, don't kill their muse, don't take their talent away from real fans. Just find things you actually enjoy for what they are.

This is why Japan is tired of western fans.

@smasithick Can't unsee that fish.

Gmail, Drive, and Meet icons have been updated.

@creepydotorg @grok I think it's about contrast. You can easily see a big grumpy guy as a threat. Now imagine a cute lady, seemingly defenseless - suddenly becoming more terrifying than a pack of rabid dogs at night.

Hey @grok why is the victim of demonic possession in movies usually a woman?

@Ferhatx458 Maybe this is a healthy version of finding a surprise onion ring in your fries.

Süpermarkete gittim ve her birinin içinde birer kiraz domates bulunan yarım marul satıyolardı. Bunun arkasındaki amacı bilen var mı?

@Sony I haven't watched the show, but that guy's facial expressions are so meme-worthy, he could easily be the next Jim Carrey.

Comment your favorite Homelander meme below 👇🏼

@OrwellNGoode Actually, that makes sense. With her own height, tall guy's genes could help with the height of her future son. Maybe it's bashing, but what are the odds of kids reaching at least the average, if both of their parents are short?

@kimmonismus Writer can use AI for: - Word lookup. - Fact-checking relevant history. - Adding realism to skills and jobs descriptions. - Brainstorming plot twists and what-ifs. - Beta reading. - Writing the whole book. As long as it's not that last one, there’s no need for scandals.

Der Spiegel, arguably Germany's best-known news outlet, has published an article that couldn't be more emblematic of the mood in Germany. Translated, it reads: "The Nobel laureate in Literature admits to using AI for her books. Is that a scandal?" Let that sink in for a second, to let the absurdity of the question sink in. A scandal is a sensational event or behavior that causes widespread public outrage and moral indignation. So the question is whether the use of AI in writing books is *in itself* scandalous, that is, grounds for moral outrage. The question of the purpose of the use, the content, isn't even raised; no criticism is even considered. Instead, the article directly asks about the value of moral condemnation simply because the use of AI is being discussed. Germany in a nutshell. Moral outrage replaces criticism. German Angst in the face of AI.

@BeckyBosch_ What's the point of moving to a country you don't even like? Especially when it's incompatible with your lifestyle and values? Life is too short to waste it on unbearable places. Even a refugee can keep looking for a better place. You don't have to stay where you are forever.

Wisst ihr was? Wenn man in ein anderes Land kommt, hat man gefälligst wenigstens den Anstand, die Kultur, die Regeln und ja auch die Esskultur .. zu respektieren. Ich hab in mehreren Ländern gelebt oder sie besucht: Peru, die USA, Deutschland, Österreich und jetzt Japan. Und in all diesen Ländern ist Schweinefleisch einfach ein ganz normaler Teil der Gastronomie. Schnitzel, knuspriger peruanischer Chicharrón, amerikanische BBQ-Ribs, österreichische Spezialitäten und ja, auch japanisches Tonkatsu… das ist für mich einfach nur ein Genuss. Und nein, nur weil es irgendwelchen Leuten aus religiösen Gründen nicht passt, werd ich jetzt sicher nicht aufhören, das zu essen. Das wär ja noch schöner. Genau DAS nenn ich egoistisch und respektlos: in ein Land kommen, alles umkrempeln wollen und dann erwarten, dass die Einheimischen plötzlich ihre eigenen Traditionen aufgeben sollen. Als Deutsche fiel es mir nicht schwer, mich an Japan anzupassen. Da wo die Mentalität zusammenpasst, funktioniert’s auch wunderbar. Die Probleme fangen doch erst dann an, wenn eben gar nichts mehr zusammenpasst und stattdessen nur noch Intoleranz herrscht. Wer ein Problem mit Schweinefleisch hat... super, dann bleibt halt zu Hause. Oder informiert euch vorher, bevor ihr in ein Land geht, in dem Schweinefleisch seit Jahrhunderten ganz normal auf den Tisch kommt. Wir passen uns nicht eurem Weltbild an. Ihr passt euch gefälligst unserem an. Ende der Diskussion.

@PolyDepression Didn't know that style had a name.

Been thinking about Vaporwave a lot lately. Not gonna lie it felt like serenity in 2016

- If Flash-Lite is the fallback, why does it deplete the quota too? - Does that mean users shouldn't waste their prompts on Flash-Lite, since it doesn't offer any benefits in terms of quota conservation? 🧵(3/3)

But I still have questions. - What happens when a user hits those limits? 1) Do all models lock out entirely? 2) Or Flash-Lite becomes the default model until the cooldown ends?🧵(2/3)

Hey, @GeminiApp, @Google, I'd really like to understand the rules. So, we have 3 models: 3.1 Flash-Lite, 3.5 Flash, and 3.1 Pro, each can use standard or extended thinking. We also have usage limits based on 5-hour and weekly windows. Easy.🧵(1/3)