Builder, Dad - YC S13 - Cofounder @PlayPokerSkill, @Cratejoy, MarketZero - Question the answers
Austin, TX
- Tweets 6,883
- Following 813
- Followers 1,322
- Likes 3,346
if openai doesn't have an answer to opus 5.5 nothing else they launch at dev day matters to me in the short term
45
yes a $500 plan will absolutely work for OpenAI and i think they can go higher. People will pay a flat rate they are comfortable with to not have to be budgeting constantly.
1
49
pacing doesn't survive first contact with a competitors model causing your users to switch because capitalism is physics
1
44
opus 5.5 makes it very difficult to want to touch any openai models right now. it's fantastic and makes the openai models feel a generation out of date and pedantic and annoying.
42
i think yesterday was actually a big inflection point in AI but it's going to take everyone a moment to realize. the quality went up yes but the usage became affordable and fast. i don't have cap anxiety for first time in a while.
44
this is because the profession will need to get rebuilt around managing outcomes and quality rather than personal code authorship. until then you just have people doing more of what they did before but sloppier and less rewarding.
I am done with this shit. It is over. The state of engineering right now is horrible. It has been half a month since I started a new role at a big company. Nobody knows anything here. The specs, code, tests, PRDs, tickets, resolution of those tickets, reports, etc., everything is made by Claude Code. Nobody on my team likes this. They are being forced to ship as much as they can. I have heard multiple times from higher management that pushing code is not a bottleneck, so why are we slow? People are working 12 to 13 hours a day just to press enter. Nobody is reading anything. Humans in corporate are doing nothing on their own. Everyone, literally everyone, from an L1 to an L7 engineer here is doing the same thing. Talk to Claude. There is no sense of victory. Nobody is resolving bugs. In reality, nobody is thinking anymore. Everything is done by LLMs. It is so soul-sucking. I would not mind it, to be honest, if we were at least given the time to check out the code and see what is going where. But no, the goal is to just ship. No matter what happens.
67
Jev is pointing out the obvious (in retrospect) thing which is the LANGUAGES themselves need primitives that allow for intelligent control flow.
45
the labs cannot conceive of a future that doesn't require their centralized decision making to thrive because they fundamentally have zero respect for the rest of humanity's ingenuity and ability to adapt. at core they believe only in their own intellect
41
Pay attention to what is happening in spiking neural networks. This fruit fly toy stuff is about to become very important. Our agents are about to have muscle memory and reflex. They break a lot of existing transformer based paradigms, but complement a neocortex well.
46
Amir Elaguizy retweeted
166,700 fruit fly neurons are flying a drone around a bedroom right now.
One camera watches an open hand. That's the entire input. The panel in the middle is the brain itself, running live, 166,700 neurons and 25 million connections, the full reconstructed wiring of a fruit fly's head.
The camera reads the hand, the brain fires, the output becomes throttle. Open palm and it climbs. Fist and it holds. Drop the hand and it comes down.
Four propellers the size of a coin, hanging off a nervous system that used to belong to an insect.
Here's the honest part. This isn't a resurrected fly. It's a computational model built from real biological wiring, bridged to a drone the builder made themselves. What nobody wrote is the flight logic. No PID tuning, no gesture presets, no code telling it how to hold altitude. They gave the neurons somewhere to send a signal, and it found the ceiling on its own.
Eleven seconds in, it's holding altitude by itself.
The fly is dead. It's still flying.
23
86
572
67,204
started using ds 4.1 flash for all my transactional work and it's honestly a pleasure to work with. It's not an astra replacement for me, it's a fantastic supplement.
49
Amir Elaguizy retweeted
I put “do not overengineer” in my AGENTS.md.
it created a lightweight extensible overengineering prevention framework
138
163
6,114
253,405
Sam and Dario get some CYA and slow down in capex
Reduction in liability?
Elon gets to see everyone else's models as part of regulatory body, which i'm sure he would like
politicians get a new boogie man to rally against
everyone wins except for startups and the US and ...
44
Amir Elaguizy retweeted
This is the thing about AI. I don't think its replacing software engineers, I think its letting software engineers build all the things they never had time for before.
Ironed out the last performance issues with my full 2.5m line codebase explorer. 120hz awesomeness. Can only upload 60fps video tho. Much nicer uncompressed
16
42
679
26,838
Amir Elaguizy retweeted
China is just going to blast past us with RSI while we self-sabotage with BS red tape that is mostly Safety theatre
🚨BREAKING: China's Z.AI has raised $5 billion.
The interesting part is where the money is going: 60% of the net proceeds will fund next-generation GLM models and what Z.AI calls its 'Fully Self Training' system.
Z.AI's stated goal is for each generation of GLM to build the environments used to train the next one, forming what the company itself calls a Recursive Self-Improvement (RSI) loop.
They just raised billions to push on it.
25
42
468
18,225