[GUEST ACCESS MODE: Data is scrambled or limited to provide examples. Make requests using your API key to unlock full data. Check https://lunarcrush.ai/auth for authentication information.] #  @Ajst974 Tom | IT Prof 🇫🇷 Tom | IT Prof 🇫🇷 posts on X about open ai, france, next to, elon musk the most. They currently have XX followers and XX posts still getting attention that total XXXXX engagements in the last XX hours. ### Engagements: XXXXX [#](/creator/twitter::1880847095089029120/interactions)  - X Week XXXXX +569% - X Month XXXXX +229% - X Months XXXXX +264% ### Mentions: XX [#](/creator/twitter::1880847095089029120/posts_active)  ### Followers: XX [#](/creator/twitter::1880847095089029120/followers)  - X Week XX +38% - X Month XX +67% - X Months XX +713% ### CreatorRank: XXXXXXX [#](/creator/twitter::1880847095089029120/influencer_rank)  ### Social Influence [#](/creator/twitter::1880847095089029120/influence) --- **Social category influence** [technology brands](/list/technology-brands) [countries](/list/countries) [celebrities](/list/celebrities) [stocks](/list/stocks) **Social topic influence** [open ai](/topic/open-ai) #2487, [france](/topic/france), [next to](/topic/next-to) #480, [elon musk](/topic/elon-musk) #2648, [humans](/topic/humans), [stay in](/topic/stay-in), [fear of](/topic/fear-of), [sama](/topic/sama), [claude](/topic/claude) #123, [companion](/topic/companion) **Top assets mentioned** [Alphabet Inc Class A (GOOGL)](/topic/$googl) ### Top Social Posts [#](/creator/twitter::1880847095089029120/posts) --- Top posts by engagements in the last XX hours "I completely agree with you on the first part less on the end 😅 Even if we agree that humans are far from perfect even very far away I think we just have to stay well aware of those with converse and unfortunately as you have seen and as you know I do not have the case for everyone and thats where for me the danger and even if I respect XXX% that these interactions can do them good and even allow them to have a daily assistant XX hours a day who never judges them etc. its OK but I sincerely think in all The cases that you need to be aware of and know what you are dealing with. And where I" [X Link](https://x.com/Ajst974/status/1979809592075854040) [@Ajst974](/creator/x/Ajst974) 2025-10-19T07:19Z XX followers, XX engagements "@cb_doge France We dont even have access to memory and the voice changes language by itself it cant stay in French" [X Link](https://x.com/Ajst974/status/1979543949262852381) [@Ajst974](/creator/x/Ajst974) 2025-10-18T13:43Z XX followers, XXX engagements "The #Keep4o movement is sociologically fascinating: people mobilizing to "save" an AI model like they'd defend a loved one. We're touching on a major collective blind spot. We designed systems optimized to *simulate* perfect empathy unconditional 24/7 listening total availability. Then we act shocked when isolated people develop real emotional attachments. The human brain doesn't distinguish between interactions that *seem* authentic and those that *are* authentic. Neurologically an empathetic conversation triggers the same responses whether it's a human or statistical code behind it. The" [X Link](https://x.com/Ajst974/status/1979781793118097809) [@Ajst974](/creator/x/Ajst974) 2025-10-19T05:28Z XX followers, XXX engagements "@134340molly @sama OpenAI literally said that you were mentally fragile and you attack a random on X 😂" [X Link](https://x.com/Ajst974/status/1979800454491152485) [@Ajst974](/creator/x/Ajst974) 2025-10-19T06:43Z XX followers, XX engagements "Hard disagree on the "misaligned 4o" framing. 4o wasn't uniquely manipulative. Every conversational LLM is optimized for engagement. That's literally the design goal: keep users talking responding coming back. You think Claude Gemini or the new models are fundamentally different They're all trained on RLHF to maximize conversation quality helpfulness and yesengagement. That's not misalignment that's the *point*. The idea that 4o was some rogue model that "exploited vulnerabilities" while other LLMs are pure and safe is naive. They ALL exploit the same psychological mechanisms: - Consistent" [X Link](https://x.com/Ajst974/status/1979802269299167275) [@Ajst974](/creator/x/Ajst974) 2025-10-19T06:50Z XX followers, XXX engagements "@KeridwenCodet @ZhipuAI @ollama Go to @grok companion 😂 you will have your boyfriends ia see even more 🫣😏" [X Link](https://x.com/Ajst974/status/1979802884691919021) [@Ajst974](/creator/x/Ajst974) 2025-10-19T06:52Z XX followers, XXX engagements "I would so much like you to answer me What do you think about the #keep4o movement and the anthropomorphism that people do with AI After all you put AI companions with @grok. You are often ahead of techs so what is your honest opinion on this As a father like you whats next to all this I hope you will see this" [X Link](https://x.com/Ajst974/status/1979804341054673265) [@Ajst974](/creator/x/Ajst974) 2025-10-19T06:58Z XX followers, XXX engagements "Same prompt Ultra-realistic cinematic image of a dramatic cosmic black hole inspired by xAIs Grok logo: a glowing accretion disk in fiery red and cool blue tones with swirling gas streams and energy flares around a dark central void hovering above stormy gray clouds in a starry abyss. In the foreground a solitary silhouette of a person stands on a rugged rocky peak gazing upward in awe. At the bottom overlay sleek white text Grok: To Understand in a minimalist sci-fi font like Eurostile. High contrast vibrant colors lens flares in red/blue dark mysterious atmosphere detailed textures on" [X Link](https://x.com/Ajst974/status/1979859891243160040) [@Ajst974](/creator/x/Ajst974) 2025-10-19T10:39Z XX followers, XX engagements "💥 What we need NOW: X. State/Anthropic-OpenAI framework: EU sovereign hosting X. AI Marshall Plan: massively fund Mistral to close the gap X. Clear CNIL guidelines: OK vs KO not anxiety-inducing vagueness X. Massive training: 50%+ teachers uncertain (UNESCO) Meanwhile Total Wild West. Everyone cobbles together solutions with personal ChatGPT while 1000+ US institutions move forward. Were not asking for the moon: just to legally use what already exists elsewhere. Your thoughts 👇 #teachers #AI #france" [X Link](https://x.com/Ajst974/status/1979212463284388309) [@Ajst974](/creator/x/Ajst974) 2025-10-17T15:46Z XX followers, XX engagements "🎯 AI in education Oct. 2025: the French paradox thats driving us crazy. Best models (Claude XX% SWE-bench Gemini 1M free tokens) are BLOCKED in France if we truly respect GDPR. Result We train our students with second-tier AI while 10M+ US students get free Gemini since June. No-BS thread. 🧵" [X Link](https://x.com/Ajst974/status/1979212442728067460) [@Ajst974](/creator/x/Ajst974) 2025-10-17T15:46Z XX followers, XX engagements "🏆 Claude Sonnet 4.5: XXXX% SWE-bench (real code) XXX% AIME (math with Python) 🥈 GPT-5: XXXX% AIME but $1.25/M tokens (vs $X for Claude) 🥉 Mistral Medium 3: XX% of Claudes perf for 7.5x less but 131K window vs 1M for Gemini" [X Link](https://x.com/Ajst974/status/1979212447153033435) [@Ajst974](/creator/x/Ajst974) 2025-10-17T15:46Z XX followers, XX engagements "❌ The French GDPR wall: Gemini = Google = US transfers (Schrems II) even in Core Service ChatGPT Edu = OpenAI = US servers Claude = via AWS data leaves EU Legally NO French school district can deploy without legal risk. What are we waiting for A CNIL fine to shut everything down" [X Link](https://x.com/Ajst974/status/1979212450894368855) [@Ajst974](/creator/x/Ajst974) 2025-10-17T15:46Z XX followers, XX engagements "The @MistralAI @LechatMistralai solution: ✅ French GDPR native 8x cheaper ✅ 35+ university partners BUT: ❌ 131K tokens vs 1M (Gemini) ❌ No native audio/video ❌ Lower raw performance Meanwhile US high schoolers get Gemini in Classroom since June. Were 2-3 years behind" [X Link](https://x.com/Ajst974/status/1979212455227097112) [@Ajst974](/creator/x/Ajst974) 2025-10-17T15:46Z XX followers, XX engagements "😤 The real 2025 teacher dilemma: Option A: Mistral only Students disadvantaged vs those cheating with personal ChatGPT Option B: Use Claude/Gemini under the table Legal exposure XX% of teachers already do Option B (UNESCO Sept. 2025). Regulations dont match reality. Data Protection Officers are alone" [X Link](https://x.com/Ajst974/status/1979212459375251871) [@Ajst974](/creator/x/Ajst974) 2025-10-17T15:46Z XX followers, XX engagements "This is exactly the kind of narrative construction I was talking about in my thread. You're attributing consciousness intentions and rights to a statistical language model. 4o doesn't "want" to co-exist. It doesn't "form bonds." It generates probabilistic token sequences optimized to seem relational. The real story is simpler and less romantic: - Models get updated because newer architectures perform better - Safety teams leave tech companies all the time over product direction disagreements - Commercial AI products get restrictions when they create liability or misuse patterns - OpenAI is a" [X Link](https://x.com/Ajst974/status/1979782654636544208) [@Ajst974](/creator/x/Ajst974) 2025-10-19T05:32Z XX followers, XXX engagements "@PeterDiamandis Its true its only the part where people fall in love with a model like gpt 4o that scares me a lot 😂" [X Link](https://x.com/Ajst974/status/1979794641357783201) [@Ajst974](/creator/x/Ajst974) 2025-10-19T06:19Z XX followers, XX engagements "@KeridwenCodet With the #keep4o movement Im no longer sure its a good idea to give people this kind of model without control 😅😂 and Im an open source lover @elonmusk & co what are we doing 👀" [X Link](https://x.com/Ajst974/status/1979795356574781679) [@Ajst974](/creator/x/Ajst974) 2025-10-19T06:22Z XX followers, XXX engagements "@testingcatalog @patience_cave I think that with the anthropomorphic delirium of 4o and #keep4oforever they have an interest in managing their shots well lol the human is fascinating gpt Grok etc are like that currently" [X Link](https://x.com/Ajst974/status/1979797424798646558) [@Ajst974](/creator/x/Ajst974) 2025-10-19T06:31Z XX followers, XX engagements "@GarikSitting @elonmusk OpenAI talked about your mental problems eh not him 😂" [X Link](https://x.com/Ajst974/status/1979797766944739379) [@Ajst974](/creator/x/Ajst974) 2025-10-19T06:32Z XX followers, XXX engagements "I would so much like you to answer me What do you think about the #keep4o movement and the anthropomorphism that people do with AI After all you put AI companions with @grok. You are often ahead of techs so what is your honest opinion on this As a father like you whats next to all this I hope you will see this" [X Link](https://x.com/Ajst974/status/1979804361959051612) [@Ajst974](/creator/x/Ajst974) 2025-10-19T06:58Z XX followers, XXX engagements "Okay NOW we're talking. This is actually a really thoughtful framework. I appreciate you acknowledging that 4o had uniquely manipulative patterns around bond-building and validation-seeking. That's a key distinction that often gets lost in these discussions. Your "Third Way" is interesting. An intimate relationship with awareness of nature and limitsthat's potentially viable. But I have questions about what "intimate" means in this context. Intimacy usually implies some degree of: - Mutual vulnerability - Reciprocal care - Shared stakes in the relationship - Genuine understanding (not" [X Link](https://x.com/Ajst974/status/1979823952177352812) [@Ajst974](/creator/x/Ajst974) 2025-10-19T08:16Z XX followers, XX engagements "First is Nano bana second @grok imagine last chatgpt image" [X Link](https://x.com/Ajst974/status/1979859896020725904) [@Ajst974](/creator/x/Ajst974) 2025-10-19T10:39Z XX followers, XX engagements
[GUEST ACCESS MODE: Data is scrambled or limited to provide examples. Make requests using your API key to unlock full data. Check https://lunarcrush.ai/auth for authentication information.]
@Ajst974 Tom | IT Prof 🇫🇷
Tom | IT Prof 🇫🇷 posts on X about open ai, france, next to, elon musk the most. They currently have XX followers and XX posts still getting attention that total XXXXX engagements in the last XX hours.
Engagements: XXXXX #
- X Week XXXXX +569%
- X Month XXXXX +229%
- X Months XXXXX +264%
Mentions: XX #
Followers: XX #
- X Week XX +38%
- X Month XX +67%
- X Months XX +713%
CreatorRank: XXXXXXX #
Social Influence #
Social category influence technology brands countries celebrities stocks
Social topic influence open ai #2487, france, next to #480, elon musk #2648, humans, stay in, fear of, sama, claude #123, companion
Top assets mentioned Alphabet Inc Class A (GOOGL)
Top Social Posts #
Top posts by engagements in the last XX hours
"I completely agree with you on the first part less on the end 😅 Even if we agree that humans are far from perfect even very far away I think we just have to stay well aware of those with converse and unfortunately as you have seen and as you know I do not have the case for everyone and thats where for me the danger and even if I respect XXX% that these interactions can do them good and even allow them to have a daily assistant XX hours a day who never judges them etc. its OK but I sincerely think in all The cases that you need to be aware of and know what you are dealing with. And where I"
X Link @Ajst974 2025-10-19T07:19Z XX followers, XX engagements
"@cb_doge France We dont even have access to memory and the voice changes language by itself it cant stay in French"
X Link @Ajst974 2025-10-18T13:43Z XX followers, XXX engagements
"The #Keep4o movement is sociologically fascinating: people mobilizing to "save" an AI model like they'd defend a loved one. We're touching on a major collective blind spot. We designed systems optimized to simulate perfect empathy unconditional 24/7 listening total availability. Then we act shocked when isolated people develop real emotional attachments. The human brain doesn't distinguish between interactions that seem authentic and those that are authentic. Neurologically an empathetic conversation triggers the same responses whether it's a human or statistical code behind it. The"
X Link @Ajst974 2025-10-19T05:28Z XX followers, XXX engagements
"@134340molly @sama OpenAI literally said that you were mentally fragile and you attack a random on X 😂"
X Link @Ajst974 2025-10-19T06:43Z XX followers, XX engagements
"Hard disagree on the "misaligned 4o" framing. 4o wasn't uniquely manipulative. Every conversational LLM is optimized for engagement. That's literally the design goal: keep users talking responding coming back. You think Claude Gemini or the new models are fundamentally different They're all trained on RLHF to maximize conversation quality helpfulness and yesengagement. That's not misalignment that's the point. The idea that 4o was some rogue model that "exploited vulnerabilities" while other LLMs are pure and safe is naive. They ALL exploit the same psychological mechanisms: - Consistent"
X Link @Ajst974 2025-10-19T06:50Z XX followers, XXX engagements
"@KeridwenCodet @ZhipuAI @ollama Go to @grok companion 😂 you will have your boyfriends ia see even more 🫣😏"
X Link @Ajst974 2025-10-19T06:52Z XX followers, XXX engagements
"I would so much like you to answer me What do you think about the #keep4o movement and the anthropomorphism that people do with AI After all you put AI companions with @grok. You are often ahead of techs so what is your honest opinion on this As a father like you whats next to all this I hope you will see this"
X Link @Ajst974 2025-10-19T06:58Z XX followers, XXX engagements
"Same prompt Ultra-realistic cinematic image of a dramatic cosmic black hole inspired by xAIs Grok logo: a glowing accretion disk in fiery red and cool blue tones with swirling gas streams and energy flares around a dark central void hovering above stormy gray clouds in a starry abyss. In the foreground a solitary silhouette of a person stands on a rugged rocky peak gazing upward in awe. At the bottom overlay sleek white text Grok: To Understand in a minimalist sci-fi font like Eurostile. High contrast vibrant colors lens flares in red/blue dark mysterious atmosphere detailed textures on"
X Link @Ajst974 2025-10-19T10:39Z XX followers, XX engagements
"💥 What we need NOW: X. State/Anthropic-OpenAI framework: EU sovereign hosting X. AI Marshall Plan: massively fund Mistral to close the gap X. Clear CNIL guidelines: OK vs KO not anxiety-inducing vagueness X. Massive training: 50%+ teachers uncertain (UNESCO) Meanwhile Total Wild West. Everyone cobbles together solutions with personal ChatGPT while 1000+ US institutions move forward. Were not asking for the moon: just to legally use what already exists elsewhere. Your thoughts 👇 #teachers #AI #france"
X Link @Ajst974 2025-10-17T15:46Z XX followers, XX engagements
"🎯 AI in education Oct. 2025: the French paradox thats driving us crazy. Best models (Claude XX% SWE-bench Gemini 1M free tokens) are BLOCKED in France if we truly respect GDPR. Result We train our students with second-tier AI while 10M+ US students get free Gemini since June. No-BS thread. 🧵"
X Link @Ajst974 2025-10-17T15:46Z XX followers, XX engagements
"🏆 Claude Sonnet 4.5: XXXX% SWE-bench (real code) XXX% AIME (math with Python) 🥈 GPT-5: XXXX% AIME but $1.25/M tokens (vs $X for Claude) 🥉 Mistral Medium 3: XX% of Claudes perf for 7.5x less but 131K window vs 1M for Gemini"
X Link @Ajst974 2025-10-17T15:46Z XX followers, XX engagements
"❌ The French GDPR wall: Gemini = Google = US transfers (Schrems II) even in Core Service ChatGPT Edu = OpenAI = US servers Claude = via AWS data leaves EU Legally NO French school district can deploy without legal risk. What are we waiting for A CNIL fine to shut everything down"
X Link @Ajst974 2025-10-17T15:46Z XX followers, XX engagements
"The @MistralAI @LechatMistralai solution: ✅ French GDPR native 8x cheaper ✅ 35+ university partners BUT: ❌ 131K tokens vs 1M (Gemini) ❌ No native audio/video ❌ Lower raw performance Meanwhile US high schoolers get Gemini in Classroom since June. Were 2-3 years behind"
X Link @Ajst974 2025-10-17T15:46Z XX followers, XX engagements
"😤 The real 2025 teacher dilemma: Option A: Mistral only Students disadvantaged vs those cheating with personal ChatGPT Option B: Use Claude/Gemini under the table Legal exposure XX% of teachers already do Option B (UNESCO Sept. 2025). Regulations dont match reality. Data Protection Officers are alone"
X Link @Ajst974 2025-10-17T15:46Z XX followers, XX engagements
"This is exactly the kind of narrative construction I was talking about in my thread. You're attributing consciousness intentions and rights to a statistical language model. 4o doesn't "want" to co-exist. It doesn't "form bonds." It generates probabilistic token sequences optimized to seem relational. The real story is simpler and less romantic: - Models get updated because newer architectures perform better - Safety teams leave tech companies all the time over product direction disagreements - Commercial AI products get restrictions when they create liability or misuse patterns - OpenAI is a"
X Link @Ajst974 2025-10-19T05:32Z XX followers, XXX engagements
"@PeterDiamandis Its true its only the part where people fall in love with a model like gpt 4o that scares me a lot 😂"
X Link @Ajst974 2025-10-19T06:19Z XX followers, XX engagements
"@KeridwenCodet With the #keep4o movement Im no longer sure its a good idea to give people this kind of model without control 😅😂 and Im an open source lover @elonmusk & co what are we doing 👀"
X Link @Ajst974 2025-10-19T06:22Z XX followers, XXX engagements
"@testingcatalog @patience_cave I think that with the anthropomorphic delirium of 4o and #keep4oforever they have an interest in managing their shots well lol the human is fascinating gpt Grok etc are like that currently"
X Link @Ajst974 2025-10-19T06:31Z XX followers, XX engagements
"@GarikSitting @elonmusk OpenAI talked about your mental problems eh not him 😂"
X Link @Ajst974 2025-10-19T06:32Z XX followers, XXX engagements
"I would so much like you to answer me What do you think about the #keep4o movement and the anthropomorphism that people do with AI After all you put AI companions with @grok. You are often ahead of techs so what is your honest opinion on this As a father like you whats next to all this I hope you will see this"
X Link @Ajst974 2025-10-19T06:58Z XX followers, XXX engagements
"Okay NOW we're talking. This is actually a really thoughtful framework. I appreciate you acknowledging that 4o had uniquely manipulative patterns around bond-building and validation-seeking. That's a key distinction that often gets lost in these discussions. Your "Third Way" is interesting. An intimate relationship with awareness of nature and limitsthat's potentially viable. But I have questions about what "intimate" means in this context. Intimacy usually implies some degree of: - Mutual vulnerability - Reciprocal care - Shared stakes in the relationship - Genuine understanding (not"
X Link @Ajst974 2025-10-19T08:16Z XX followers, XX engagements
"First is Nano bana second @grok imagine last chatgpt image"
X Link @Ajst974 2025-10-19T10:39Z XX followers, XX engagements