Haider.
@slow_developer
together, we build an intelligent future.
قد يعجبك
Wow, this is massive and unbelievable. 50 subscribers. i’m super thankful to all of you, seriously. it feels like i’m just getting started, and there’s more fun on the way. btw, those video clips you see on my feed every day? yes, they were edited by my junior (i paid him) so…
this setup is working well if you like switching models: > gpt-5-high/codex for debugging and planning > gpt-5-md for implementation > sonnet 4.5 for refactoring and documentation > gemini 2.5 pro for web and document searches
UBI is a bad idea, in my view it will only be taken seriously when advanced AI visibly replaces jobs right now, there's no real political push most ppl don’t know about LLMs, and those who do rarely use them seriously until AI displaces frontline roles at scale, political…
Sam Altman says AI may need a new hardware form factor OpenAI and Jony Ive are exploring a family of devices to make AI easier to use — this will take time (not this year, maybe not next) The Goal is a new kind of computer built around AI, a companion through your life
GPT-5-high = top tier for agentic work, coding, and research GPT-4.5 = unmatched for creative writing and emotional intelligence most people hating on GPT-5 were using the free tier for casual chat; they haven't tried what GPT-5 with thinking via API can actually do
still don't get why some people think LLMs have stopped improving that doesn't add up. we need benchmarks to know which model is better -- without them, you can't tell whether one model is actually ahead if we can't tell when something is smarter than us, truth turns into…
there's a lot of misunderstanding about AI people call it "just autocomplete", but that misses the point these systems show capabilities far beyond those of the average human recently, gemini 2.5 DeepThink and gpt-5 won gold in the ICPC contest far beyond autocomplete? maybe
it's crazy how fast programming has changed this year in 2023, we went from autocomplete tools like Copilot to AI Agents that do 90% of the work and now we spend the day talking to Codex or Claude Code, reviewing their outputs, with at most 10% actual coding from us
Sam Altman says AI will blur the gap between passive watching and active creating Instead of just leaning back for a movie or sweating through a game, you'll engage mid-way — shaping what unfolds Sora shows the shift: most users are now creating, not just watching
we are still far from RSI AI that can redesign and improve itself even though current systems can think for hours, solve IMO/ICPC problems, win gold medals, and outperform top humans but once that stage is reached, AI could quickly advance to AGI and beyond in a short time
just like GPT-5 has pushed the basics forward Gemini 3 won't have to do much just avoid doing anything unnecessary. focus on smart, practical improvements > reduce hallucinations > more reliable tool calling > expand the context window > add some agent-like functionality
ignore the doomers start integrating AI into your life AI doesn't care if you love it or hate it no technology has grown this fast or attracted this much funding its impact on every industry is inevitable start late, watch others win. the choice is yours
we saw news a year ago that Elon was building an AI to be the best in the world we doubted it, but now they've built a cutting-edge lab in just 2 years in about a year, xAI launched: 3 base models open-source model separate coding model image-and-video model low-cost…
GPT-5 Pro is a beast in math it set a new record on FrontierMath Tier 4 with 13% solving 6/48 problems, including one that no other model has solved tested twice, it reached a combined pass@2 of 17% - Grok 4 Heavy: 2-3% - Gemini 2.5 DeepThink: 12%, with one problem less
Gemini 2.5 DeepThink sets a new bar on FrontierMath EpochAI reports: - 29% on Tiers 1-3, 10% on Tier 4 - tested in the gemini app (public IMO-style model) with code + web tools - IMO checks: 61% in 2025 (bronze level) but misses two harder 2024 problems needing long, creative…
sam was right. for most people, AI is just chatGPT meanwhile, we already have systems that already match top humans and win hard contests to be fair, most folks use the free GPT-4o, which will confidently tell you that: "hey, i'm really dumb, but at least i'm free"
United States الاتجاهات
- 1. Jets 70.7K posts
- 2. Jets 70.7K posts
- 3. Aaron Glenn 3,552 posts
- 4. Justin Fields 5,265 posts
- 5. #HardRockBet 2,051 posts
- 6. Garrett Wilson 2,439 posts
- 7. HAPPY BIRTHDAY JIMIN 93.5K posts
- 8. London 204K posts
- 9. #OurMuseJimin 135K posts
- 10. #JetUp 1,564 posts
- 11. Peart 1,730 posts
- 12. Tyrod N/A
- 13. #DENvsNYJ 1,661 posts
- 14. Bo Nix 2,233 posts
- 15. #30YearsofLove 125K posts
- 16. Bam Knight N/A
- 17. Sherwood 1,168 posts
- 18. Sean Payton N/A
- 19. Hail Mary 2,453 posts
- 20. Kurt Warner N/A
قد يعجبك
-
Alamin
@iam_chonchol -
David Mráz
@davidm_ml -
Muzzammil
@DevMuzzammil -
Ajay Sharma
@ajaysharma_here -
Mujeeb Ahmed
@hey_mujeebahmed -
Anas
@muhamdanaskhan -
Amena | Figma Expert
@AmenaiSabuwala -
Ali Sufian
@aliscodes -
D-Coder
@Damn_coder -
StackAI
@StackAI_HQ -
Pradeep🕵️♂️
@pradeepeth -
Ahsan Noor
@ahsannoor101 -
Pritesh Kiri
@PriteshKiri -
Angry Tom
@AngryTomtweets -
Ali Reza
@alirdev
Something went wrong.
Something went wrong.