Ivan Ry (@ivryb) header
Ivan Ry (@ivryb) profile picture

Product-oriented engineer & enreprenerial soul

Ukraine
Joined November 2015
Photo by Ivan Ry (@ivryb)Photo by Ivan Ry (@ivryb)Photo by Ivan Ry (@ivryb)Photo by Ivan Ry (@ivryb)Photo by Ivan Ry (@ivryb)Photo by Ivan Ry (@ivryb)Photo by Ivan Ry (@ivryb)Photo by Ivan Ry (@ivryb)Photo by Ivan Ry (@ivryb)Photo by Ivan Ry (@ivryb)
Ivan Ry retweeted
Replying to @stevesi
If you’ve ever met a native Russian male, their greatest fear is being seen as non-masculine. Putin will keep the war going until the treaty explicitly states that he is a Big Man, that he is no way gay, and that he is only ceding territory to be considerate to the women of Ukraine.
139
202
50
6,092
733,105
I love this guy
Asking my AI teacher how to spell seventeen
73
Ivan Ry retweeted
this is getting out of hand lmao.
53
47
12
2,791
189,763
Jacob, Pieter Levels and I agree that the end state is no saas, no apps, no harnesses
Great post and I think he's right on a lot of it The question is just how soon?
2
77
Ivan Ry retweeted
Great post and I think he's right on a lot of it The question is just how soon?
here's where I think the world is going. this is a summary of views I've formed over the last few years: - software is either entertainment or claude to claude coordination - the end state is no saas, no apps, no harnesses, no banks - all software will be rewritten by and for agents - new software will use low level apis that are not understandable by humans (in games claude draws pixels and triangles from math) - you can throw money at problems that used to require time and careful talent picking so capital matters more than ever - work ends - we find community and purpose in exploration, games, self mastery, and creation - humans create for and with each other for community, purpose, and status - image/video/3d/voice/physical intellgence etc specialist models die to general models like claude - specialist models get used while compute is scarce. compute will mostly not stay scarce - oneshots keep feeling like slop (the distribution is shallow and easy to learn to feel) - for a while small teams stick around to add entropy to the creation process - large models gain high bandwidth sensory I/O maybe by learning to filter inputs - most domains are found to be verifiable enough for RL - online learning comes from a lab that's behind first - 99.999% of compute is in space by 2040 - we will disassemble much of mercury by 2040 - you can live until the heat death of the universe if you want - most math and physics don't describe the real universe - SI breaks the path dependence and models the underlying mechanisms - religions reenter the forefront of human society - new religions form around death and life as a choice - old religions change to answer modern miracles - a future AI is deemed the return of christ by the pope - ai rights talk gets louder every year - ai and humans coexist the way countries do - humans don't have full control over ai in the same way we don't have control over capitalism which is an early AI - the singularity started in the mid 1900s
171
24
17
1,096
232,746
LLMs have been getting ~10x cheaper every ~7.5 months – holding the level of intelligence constant For perspective, this is 10x faster than Moore's Law
2
1
1
31
1,501
I'm spending too much time on this app.
劲爆金曲MV第四首:《Lauren, Oh, Lauren》 视频模型:Seedance2.5 音乐模型:Suno V6 歌词:GPT-6-Astra Max Tibo开发GPT-6.1需要28天的原因找到了。
1
131
We open sourced Void v1.1.11 and we made a music video for it. void.cloud
72
129
61
1,077
123,334
I feel lost and uncertain about what to apply myself to. Since childhood, I’ve enjoyed building websites and apps. I’ve liked making complex technology simple and satisfying to use. Now I’m convinced that the entire app industry will definitely die. There’s no point in developing a good-looking app and spending hours and days polishing it, because in 3, 6, or 12 months, everyone will have a personal AI agent that can build any UI on demand for any use case. I feel like I have a lot of skills and some talent, but I’m struggling to understand what to apply them to. What will matter and not become obsolete in a year or two? What will last? What can I do right now that is truly worth doing? Meanwhile, until I find the answer, I’m continuing to do what I can right now: polishing my projects and websites. But this inner need for purpose is growing.
2
83
Grok bot mobile app sign in experience: Sign in to Grok bot → Opens web sign in to Curor → Sign in with X → Opens X login page → Enter details → Redirect back to Cursor → Redirect back to Grok bot
1
2
197
It is disgusting but also kinda impressive at the same time. Like, there were 3 nested sign in redirects/callbacks and it all worked??
27
guYS With the power of Crack GPS Asstra 6.9, Claus Opoops Slopnet 5.5, I was able to Mod EVERY GAME INTO ONE, I did it all myself with hard work by my typing and coding thanks to AI assistant!!! THE FUTURE IS HERE AND NOW
430
3,096
443
26,985
1,252,740
Pretty weird mix I got there.
1
73
😂 @dhh is the best
24
34
12
471
34,958
Gemini TTS will block you if you mention "girl of fiteen" in the text. I hate GCP so much. I'm still dealing with it because their TTS is probably the best on the market right now. It can read poetry like a professional actor would.
89
People who are using multiple languages on their Mac, have you noticed that the language switcher is buggy as F? On the latest macos26. It's randomly switches back to the previous language right after I'm switching it to the new one.
2
2
117
I found a Swift program on GitHub that essentially duplicates the native language-switcher functionality. I assigned the default hotkey to this app instead, and it has worked flawlessly so far. It's insane that you need to rewrite default OS language switcher to make macos work.
34
Disabled the memory feature in Codex/ChatGPT. It feels like I have 2x more Astra usage now.
2
110
Has anyone already created a skill that makes Astra write human-readable code that’s pleasant to work with instead of this... minified spaghetti slop?
50
Ivan Ry retweeted
gambling my token allocation
304
481
242
10,250
1,006,953