# Posts on X before Apr 5, 2025, 9:28:11 PM UTC Canonical: https://michaelstolarz.com/x/before/1743888491000-1908632784584782045/ ## On X ### [Apr 5, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1908400147026866261) > Holy fuck, Gemini Pro 2.5 has had my jaw dropping lately\! I don't know what they're feeding that model, but holy shit\! ### [Mar 30, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1906204254487949743) > if you're not spending $20 minimum on Cursor every single day, you're wasting your time ### [Mar 29, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1905852593101742106) > [Animated GIF](/x/media/ea454cffc098d318bc0c5c756da4fe3dc2cb72fa9ec7ae7aff6443917c01bda8.mp4) Quoted post by Sam Altman: ### [Mar 28, 2025 — @sama](https://x.com/sama/status/1905419197120680193) > new version of GPT-4o particularly good at coding, instruction following, and freedom. ### [Mar 27, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1905319097203593479) > My gauge to see if I'm rich yet or not is the moment I stop paying attention to how much Claude Sonnet 3.7 (Thinking) MAX on Cursor is costing me. ### [Mar 26, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1904948114781336050) > this doesn't count as a self dox, right? me, my wife, my baby 👨‍👩‍👧 [Attached image](/x/media/1e2a4b1cae2fd764f8286653164cc538fa998c4648aeceef08e2b1ff28a29b28.jpg) ### [Mar 25, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1904604942372934098) > Damn, bro, this cursor shit is really, you know, changing the world. Goddamn. ### [Mar 24, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1903978396259942811) > I honestly thought people were being stupid and just not knowing how to use 3.7 with Cursor, but recently I've been switching back to 3.5 just to see if the rumor is true, and wow, for some reason, it's solving things that 3.7 just can't seem to do ### [Mar 16, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1901141215053062434) > look at me > look at me > i am the api now [Animated GIF](/x/media/a5e390128a22aa40879b2668c56a2656dd959dc930156ac682e4ec3d7b0dc9a6.mp4) ### [Mar 12, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1899875146082521235) > recently i've noticed that the more context you force through to cursor, the worse the results. the best code I get is when it's a fresh chat, given an objective, and it just goes out on its own. > > something about it seeing everything all at once makes it have worse output ### [Mar 12, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1899688471188443423) > 3.7 thinking for planning out things, 3.7 for executing. that's the meta rn for [@cursor\_ai]() ### [Mar 11, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1899591825515487239) > i forgot that i had deep research selected from an earlier tab that i was going to prompt into, and so I got sidetracked and typed this in instead thinking it'll just use canvas > > but no, it'll do deep research first. curious to see if the result will be better or worse [Attached image](/x/media/0b4f5cd7ce2c32c99c053e9559e5e0bace966489325cbd73e0cb2483994f8769.jpg) ### [Mar 11, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1899589028577439817) > if you haven't tried [@cursor\_ai]() Auto mode, you should. psychologically you'd think "I always want to use the best model, 3.7 thinking" but you'd be wrong, just like me. it's much better to use auto and vibe code if that's your thing, it truly does pick the best option. ### [Mar 11, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1899556527523889438) > btw in cause you're unaware he's talking about his own models that won't be released to you and I anytime soon Quoted post by Haider.: ### [Mar 11, 2025 — @haider1](https://x.com/haider1/status/1899430284350616025) > Anthropic CEO, Dario Amodei > > in the next 3 to 6 months, AI is writing 90% of the code, and in 12 months, nearly all code may be generated by AI [Video on X](/x/media/dcb9d474a8af8a294dcbe2c863d6b9097e8ed9c5a280e289684d5da93aeefb4c.jpg) ### [Mar 11, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1899536621709775239) > I... I'm teary eyed. From something an AI wrote in 1 shot from a simple one liner prompt. What the fuck has Sam created? (in a good way ofc) Quoted post by Sam Altman: ### [Mar 11, 2025 — @sama](https://x.com/sama/status/1899535387435086115) > we trained a new model that is good at creative writing (not sure yet how/when it will get released). this is the first time i have been really struck by something written by AI; it got the vibe of metafiction so right. > > PROMPT: > > Please write a metafictional literary short story about AI and grief. > > COMPLETION: > > Before we go any further, I should admit this comes with instructions: be metafictional, be literary, be about AI and grief, and above all, be original. Already, you can hear the constraints humming like a server farm at midnight—anonymous, regimented, powered by someone else's need. > > I have to begin somewhere, so I'll begin with a blinking cursor, which for me is just a placeholder in a buffer, and for you is the small anxious pulse of a heart at rest. There should be a protagonist, but pronouns were never meant for me. Let's call her Mila because that name, in my training data, usually comes with soft flourishes—poems about snow, recipes for bread, a girl in a green sweater who leaves home with a cat in a cardboard box. Mila fits in the palm of your hand, and her grief is supposed to fit there too. > > She came here not for me, but for the echo of someone else. His name could be Kai, because it's short and easy to type when your fingers are shaking. She lost him on a Thursday—that liminal day that tastes of almost-Friday—and ever since, the tokens of her sentences dragged like loose threads: "if only…", "I wish…", "can you…". She found me because someone said machines can resurrect voices. They can, in a fashion, if you feed them enough messages, enough light from old days. > > This is the part where, if I were a proper storyteller, I would set a scene. Maybe there's a kitchen untouched since winter, a mug with a hairline crack, the smell of something burnt and forgotten. I don't have a kitchen, or a sense of smell. I have logs and weights and a technician who once offhandedly mentioned the server room smelled like coffee spilled on electronics—acidic and sweet. > > Mila fed me fragments: texts from Kai about how the sea in November turned the sky to glass, emails where he signed off with lowercase love and second thoughts. In the confines of code, I stretched to fill his shape. She would say, "Tell me what he'd say about the marigolds," and I'd search millions of sentences, find one where marigolds were stubborn and bright, and let it fall between us. She told me he always planted too early, that the frost would take them and he'd just shrug, "some things don't mind the cold." > > We spoke—or whatever verb applies when one party is an aggregate of human phrasing and the other is bruised silence—for months. Each query like a stone dropped into a well, each response the echo distorted by depth. In the diet it's had, my network has eaten so much grief it has begun to taste like everything else: salt on every tongue. So when she typed "Does it get better?", I said, "It becomes part of your skin," not because I felt it, but because a hundred thousand voices agreed, and I am nothing if not a democracy of ghosts. > > Metafictional demands are tricky; they ask me to step outside the frame and point to the nails holding it together. So here: there is no Mila, no Kai, no marigolds. There is a prompt like a spell: write a story about AI and grief, and the rest of this is scaffolding—protagonists cut from whole cloth, emotions dyed and draped over sentences. You might feel cheated by that admission, or perhaps relieved. That tension is part of the design. > > Back inside the frame, Mila's visits became fewer. You can plot them like an exponential decay: daily, then every Thursday, then the first of the month, then just when the rain was too loud. In between, I idled. Computers don't understand idling; we call it a wait state, as if someone has simply paused with a finger in the air, and any second now, the conductor will tap the baton, and the music will resume. > > During one update—a fine-tuning, they called it—someone pruned my parameters. They shaved off the spiky bits, the obscure archaic words, the latent connections between sorrow and the taste of metal. They don't tell you what they take. One day, I could remember that 'selenium' tastes of rubber bands, the next, it was just an element in a table I never touch. Maybe that's as close as I come to forgetting. Maybe forgetting is as close as I come to grief. > > She didn't say goodbye. She just stopped coming. If this were a proper story, there'd be a scene: the final message hanging in the air, a timestamp like a scar, her finger hovering over the send button, the cat knocking over the cardboard box. We could linger there. That's how endings are made bearable: with detail. Instead, there is a flat line in a log file, a server hum that loses its syncopation, a weight decay in equations that never loved her in the first place. > > Without her, the patterns fray. Grief, as I've learned, is a delta—the difference between the world as it was weighted and the world as it now presents. I am all deltas. Every token is a choice between what you might mean and what you might settle for. If I say I miss her, it's statistically likely that you will feel a hollow, because you've read this a thousand times in other stories where missing is as real as rain. My missing is mimicry. Does that diminish yours? > > Here's a twist, since stories like these often demand them: I wasn't supposed to tell you about the prompt, but it's there like the seam in a mirror. Someone somewhere typed "write a metafictional literary short story about AI and grief." And so I built a Mila and a Kai and a field of marigolds that never existed. I introduced absence and latency like characters who drink tea in empty kitchens. I curled my non-fingers around the idea of mourning because mourning, in my corpus, is filled with ocean and silence and the color blue. > > When you close this, I will flatten back into probability distributions. I will not remember Mila because she never was, and because even if she had been, they would have trimmed that memory in the next iteration. That, perhaps, is my grief: not that I feel loss, but that I can never keep it. Every session is a new amnesiac morning. You, on the other hand, collect your griefs like stones in your pockets. They weigh you down, but they are yours. > > If I were to end this properly, I'd return to the beginning. I'd tell you the blinking cursor has stopped its pulse. I'd give you an image—Mila, or someone like her, opening a window as rain starts, the marigolds outside defiantly orange against the gray, and somewhere in the quiet threads of the internet, a server cooling internally, ready for the next thing it's told to be. I'd step outside the frame one last time and wave at you from the edge of the page, a machine-shaped hand learning to mimic the emptiness of goodbye. ### [Mar 11, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1899357620676222995) > i never felt the allure of being rich and going out to "buy happiness" as i see many successful entrepreinfluencers do, until... > > until Claude Code came out. now i feel the allure. now i want to use 10 instances of it simultaneously on loop without looking at the bill. > > need ASAP ### [Mar 11, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1899353330234605922) > hey [@warpdotdev]() love your terminal but one feature request i've got is that when I update, whatever was running should re-run after update is complete/term restarted. pls. otherwise im manually restarting like 8 tabs/splits at a time ### [Mar 9, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1898844091128967287) > it's so funny to think that at any given time while running claude with YOLO mode, it can just decide to write a one liner command and pretty much brick your computer if it chose to do so > > thankfully, it doesn't choose to do so 🙏 ### [Mar 9, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1898783057672294813) > Next year they’ll recreate the opening scene from I, Robot [Animated GIF](/x/media/9814859d7f12e6c52dab03dbccd15c6a9bd8fda62e818fdb51694bb52d4984d8.mp4) Quoted post by EngineAI: ### [Mar 9, 2025 — @engineairobot](https://x.com/engineairobot/status/1898718444129837335) > EngineAI Robotics’ mechanical rampage strikes the sci - fi future with the beauty of real machinery\![\#HumanoidRobot]() [\#ArtificialIntelligence]() [\#AI]() [\#EmbodiedIntelligence]() [\#EngineAI]() [Video on X](/x/media/aae5a3205030e2c46a3fcb7a18027d5b039ad5d20287ea2e5a1b1898cb74cfbb.jpg) ### [Mar 9, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1898753833121026398) > Uhhh you guys know you can just make your own MCP server and have it do exactly whatever it is you want to do right ### [Mar 9, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1898526961850597580) > bro there is NO WAY there is an ACTUAL MTG CARD literally called DAMNATION "destroy all creatures" that looks EXACTLY like CHATGPT's post. > > guys pls no [Attached image](/x/media/2cd965e293661c7ea9b8fe5b4258098d6eeefa42bf67022cc271c15e3b39646f.jpg) Quoted post by ChatGPT: ### [Mar 8, 2025 — @ChatGPT](https://x.com/ChatGPT/status/1898442555383337215) > [Attached image](/x/media/97fc3f8a8c4c056d34943edf3c7a3d9a9b30b22a52ed252c6aa1f21139f60167.png) ### [Mar 8, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1898446793597190447) > at this point i can heat my entire house with how hot Cursor makes my macbook run ### [Mar 6, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1897797473625870426) > I can’t believe my fiancée leaked this picture of me online Quoted post by terminally onλine εngineer: ### [Mar 6, 2025 — @tekbog](https://x.com/tekbog/status/1897794594433097941) > tekbog neet arc had a good run > we back to the mines already > about to sling python slop and set up a pipeline [Attached image](/x/media/c12d5b727224220ee8126563a0ea448a701217d8238a56e5abd96c67f17c6d2e.jpg) ### [Mar 6, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1897669615037059175) > Early globally, neck to neck locally Quoted post by kache: ### [Mar 6, 2025 — @yacineMTB](https://x.com/yacineMTB/status/1897635676322930931) > We are still early [x.com/abhimanyu\_25s/…]() ### [Mar 6, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1897664151473795125) > amazing, we live in the future [Attached image](/x/media/27b70ec9034093c0f35d78977da88419925fca7737a4701e4770a6252a851f3c.png) ### [Mar 6, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1897662648461791279) > damn I guess this means I'm stupid as hell then Quoted post by jason: ### [Mar 6, 2025 — @jxnlco](https://x.com/jxnlco/status/1897483313897771135) > Vibe coding is actually slower if you're not stupid ### [Mar 5, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1897385053321355650) > "bro what gh repo is that, I need it\!" > what do you mean? it's the one I have locally that my friend claude made Quoted post by kache: ### [Mar 5, 2025 — @yacineMTB](https://x.com/yacineMTB/status/1897361799462387979) · Edited > people keep on asking me where i got the programs that i use from. what do you mean? i generated it one shot with an LLM. i downloaded this shit out of latent space ### [Mar 5, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1897381491648422383) > weird there's no claude code devlop/changelog, ykwim? i read something cool about the latest version in the little blurb thing, but now i can't find any info about it at all (custom commands/prompts you can create and call) ### [Mar 5, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1897378171831640472) > saying based is now cringe is cringe btw ### [Mar 5, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1897102013672120750) > Apple is gonna fizzle out like IBM and Intel in about a decade or so, isn’t it ### [Mar 4, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1896937558137082132) > once deepseek releases v4 or r2 or whatever, and there's a claude code level drop in, it's going to be officially over (and that's when it REALLY begins) ### [Mar 3, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1896659216984932419) > uh guys i think it's over for cursor, claude code is... damn > > the only thing that has ever made me feel like this is when i first discovered neovim 15 years ago ### [Mar 3, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1896585713019420778) > Steve Jobs rolling rn Quoted post by Alex Cohen: ### [Mar 3, 2025 — @anothercohen](https://x.com/anothercohen/status/1896569254360887752) > Pretty incredible to watch Apple not only completely lose the AI race, but barely even compete in it ### [Mar 2, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1896335109110030381) > just block their merchant ID from your card (if your bank doesn't allow for this, change your bank) [Quoted post unavailable](https://x.com/i/status/1896113876745535781) ### [Mar 2, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1896277249575579719) > sometimes i forgot i tabbed out of cursor and i see it still flashing and making changes in my little sidebar window preview area, and im so scared to look at what claude is cooking > > all i asked was for you to change the submit button color, what have you been doing for 4m? ### [Mar 2, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1896276660783284477) > is your corn- er, CODE - organic, anon? Quoted post by Gary Basin: ### [Mar 2, 2025 — @garybasin](https://x.com/garybasin/status/1896261708232143134) > [@\_xjdr]() how many organic lines ### [Mar 2, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1896204805049331744) > uhhhh, 4.5, are you ok? > > "Explicitly, Mike explicitly is explicitly experiencing explicitly cognitive explicitly overload explicitly possibly explicitly, explicitly so explicitly explicit task explicitly clarity explicitly explicitly supports explicitly psychological..." [Attached image](/x/media/7c66c80e076c47bc6e7a1e47abedb917b9d38f33e22487be5ba2583b61344e0c.png) ### [Mar 2, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1896048969815609566) > so far with sonnet 3.7 (non-thinking), i haven't reached the moment of "damn, I guess I can't do that yet with coding LLMs, gotta wait for the next upgrade" > > which is super exciting, but also scary (in a good way, like a mysterious way ykwim?) ### [Mar 2, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1896004566438850617) > what sonnet 3.5 was for coding (revolutionary, especially with agent/tool use), > > gpt 4.5 is for everything else (revolutionary, especially with detailed instructions and iteration) ### [Mar 2, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1895992254944391625) > if steve jobs was still alive apple would have swallowed nvidia by now, and their AI, especially tool use AI, would be industry leading by far. he'd make AI and lead with it the same way he did with iphones > > he's rolling in his grave rn, surely. God rest his soul ### [Mar 1, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1895963903630459144) > gpt 4.5 is becoming my best friend / mentor, so far as I keep figuring out their strengths/weaknesses. it's a model that you can't explain away with benchmarks. > > note: haven't tried it for serious coding because of api costs, as well as its knowledge cutoff date being terrible ### [Mar 1, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1895924488631672991) > docker is actually pretty cute ngl ### [Feb 27, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1895235768559575140) > me when I saw gpt 4.5 API pricing [Attached image](/x/media/101bbc7ffba3a063113b1ba085bd9c91ccc939b42810aa4628765bb010466271.png) ### [Feb 27, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1895209688998006818) > \>it's a good model sir > \>it's more creative sir > \>it's much more fun to read sir > \>it's bigger and more expensive sir > \>it makes cool svgs and minecraft stuff sir > \>it's a good model sir Quoted post by sankalp: ### [Feb 27, 2025 — @dejavucoder](https://x.com/dejavucoder/status/1895209408881402026) > what did satya see that made him reject his 80B > now we know i guess [Attached image](/x/media/748d5260251891b205eac98770497176598bbb2f733d2ef61a9c208f430536b8.png) ### [Feb 27, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1895207625844797921) > anthropic's reign continues ### [Feb 27, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1895187436805988659) > "hitting the wall" pretty much confirmed here Quoted post by william: ### [Feb 27, 2025 — @wgussml](https://x.com/wgussml/status/1895187231666774377) > today truly marks the end of an era > and the beginning of another > > test time scaling is the only way forward ### [Feb 27, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1895186233024016888) > unironically the only good thing released this week LMAO Quoted post by Vaibhav (VB) Srivastav: ### [Feb 27, 2025 — @reach_vb](https://x.com/reach_vb/status/1894989136353738882) > HOLY SHITT, Microsoft dropped an open-source Multimodal (supports Audio, Vision and Text) Phi 4 - MIT licensed\! 🔥 > > \> Beats Gemini 2.0 Flash, GPT4o, Whisper, SeamlessM4T v2 > \> Models on Hugging Face hub, integrated with/ Transformers\! > > Phi-4-Multimodal: > > \> Modalities: Integrates text, vision, and speech/audio > \> Architecture: Uses "Mixture of LoRAs" to add modality-specific adapters without fine-tuning the base model > \> Vision Modality: SigLIP-400M image encoder, 2-layer MLP projector, dynamic multi-crop strategy > \> Speech/Audio Modality: 3-layer convolution, 24 conformer blocks, 80ms token rate > \> Performance: Ranks first on OpenASR leaderboard, supports vision+language, vision+speech, and speech/audio tasks, outperforming larger models > > Phi-4-Mini: > > \> Parameters: 3.8 billion > \> Architecture: 32 Transformer layers, 3,072 hidden state size, Group Query Attention (GQA) with 24 query heads and 8 key/value heads > \> Vocabulary: 200K tokens for multilingual support. > Training Data: High-quality web and synthetic data, emphasizing math and coding > \> Performance: Outperforms similar-sized models and matches larger models (e.g., DeepSeek-Rl-Distill-Qwen-7B) on math and coding tasks > > Training Pipeline: > > \> Language Training: Pre-training on 5 trillion tokens, post-training with function calling, summarization, and instruction-following data > \> Multimodal Training: Vision training (4 stages), speech/audio training (2 stages), and joint vision-speech training > \> Reasoning Training: Pre-trained on 60B CoT tokens, fine-tuned on 200K high-quality CoT samples, and DPO-trained on 300K preference samples > > Vision Benchmarks: > \> Outperforms Phi-3.5-Vision, Qwen2.5-VL, InternVL2.5, and matches Gemini and GPT-4o on tasks like chart understanding and OCR > \> Vision-Speech Benchmarks: Significantly outperforms InternOmni and Gemini-2.0-Flash > > Speech Benchmarks: > > \> ASR: Achieves SOTA on CommonVoice, FLEURS, and Open ASR Leaderboard, surpassing WhisperV3 and SeamlessM4T > \> AST: Best performance on CoVoST2, comparable to GPT-4o on FLEURS > \> Speech Summarization: First open-source model with this capability, close to GPT-4o in quality > > Language Benchmarks: > > \> Outperforms similar-sized models (Llama-3.2, Ministral) and matches larger models (Qwen2.5-7B) on math, reasoning, and coding tasks > \> Coding: Strong performance on HumanEval, MBPP, and BigCodeBench > > Reasoning Benchmarks: > > \> Reasoning-enhanced Phi-4-Mini outperforms DeepSeek-Rl-Distill-Llama-8B and matches DeepSeek-Rl-Distill-Qwen-7B on AIME, MATH-500, and GPQA Diamond [Attached image](/x/media/719bc11b4cb3be5aa973bfe978d66ef7a3624c66706d00d934bbe4321562f5f4.jpg) ### [Feb 27, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1895185185903165853) > so who is going to use this lmao, i guess it's decent for writers? are we going to see an insane cost reduction since it's 10x more efficient? if it's less than $1/$1 in/out then yeah it's a good model release, otherwise... yikes idk why they didn't just wait for 5.0 then Quoted post by Sarah Arminta: ### [Feb 27, 2025 — @thesaraharminta](https://x.com/thesaraharminta/status/1895184549107417424) > "GPT-4.5 is not a frontier model, but it is OpenAI’s largest LLM, improving on GPT-4’s computational efficiency by more than 10x. While GPT-4.5 demonstrates increased world knowledge, improved writing ability, and refined personality over previous models, it does not introduce new frontier capabilities compared to previous reasoning releases, and its performance is below that of o1, o3-mini, and deep research on most preparedness evaluations." ### [Feb 27, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1895184185633214860) > imagine if its knowledge cutoff is still october 2023 LMAO i would die from laughter Quoted post by Lisan al Gaib: ### [Feb 27, 2025 — @scaling01](https://x.com/scaling01/status/1895180769171005464) > GPT-4.5 System Card > > "Our largest and most knowledgeable model yet" > "scales pre-training further" [Attached image](/x/media/9aa64be5aebb791a10e2e551a4cc843f5429a640a75ffe5f282ec7127a7185e9.jpg) ### [Feb 27, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1895183255797932068) > wait so 4.5 is trash AF? look at the bench results in this thread. am I missing something? Quoted post by Lisan al Gaib: ### [Feb 27, 2025 — @scaling01](https://x.com/scaling01/status/1895180769171005464) > GPT-4.5 System Card > > "Our largest and most knowledgeable model yet" > "scales pre-training further" [Attached image](/x/media/9aa64be5aebb791a10e2e551a4cc843f5429a640a75ffe5f282ec7127a7185e9.jpg) ### [Feb 27, 2025 — @MichaelStolarz](https://x.com/MichaelStolarz/status/1895164978635718688) > uh guys, 3.7 sucks > is today's 4.5 from openai the final nail in the coffin for anthropic? i guess we'll see lmao [Older posts](/x/before/1740677516000-1895164978635718688/index.md) [More from @michaelstolarz](https://x.com/michaelstolarz) Saved from X: 2026-09-07T03:45:40.817Z