erekhron (@erekhron) profile picture

Genreless Future Music Producer. CTO @ WeBrain AI & 1492 Better Together https://teapawt.org/t.co/cyISotGmCt

Hungary, Monor
Joined December 2014
In the age of ai... Still creating and jamming is the greatest if gifts. Serum Reese + Juno60 + jungle x dnb
1
1
367
Behind the Booth with Final Fantasy VII Revelation’s voice cast. #FF7R
331
3,326
709
18,333
1,027,788
Most exciting about jev how it unlocks evals / relevancy / semantic analysis across a whole lot of areas. I'm still on my wait list but can't wait to experience it fully.
1
27
Bro they work like a wishlamp. I started creating my own personal orchestrator for exactly this but with Projects there is no need. @cursor_ai *slow clapping*
Introducing Projects, a new way of working in Cursor. Rather than creating a chat for every task, you work with a coordinator agent in a single, persistent thread. Like @bot, your agent is always on, proactively manages work with subagents, and improves over time.
1
26
erekhron retweeted
China published the most uncomfortable paper on vibe coding. ETH Zurich tested 100 developers in a controlled, commercial-grade vibe coding environment to see who actually succeeds. The findings are brutal. The researchers tracked computer science achievement, written communication skills, and general cognitive reasoning. They wanted to see what actually predicts vibe coding proficiency when you never touch a line of source code yourself. Two major predictors emerged. Written communication proficiency mattered. The ability to structure thoughts and articulate intent unambiguously in text directly impacts what the AI builds. But that wasn't even the main takeaway. Computer science achievement was a massive, dominant predictor of success. Even when researchers controlled for general intelligence and reasoning skills, CS background still heavily dictated who built working software and who completely crashed. In fact, CS knowledge contributed roughly twice the unique predictive variance of writing skills alone. Why? Because vibe coding isn't about writing code. It’s about debugging logic. When an AI agent builds a complex application and quietly breaks an edge case under the hood, a non-technical user looks at the glowing UI and assumes it works. They don't know what questions to ask. They don't know what logic to challenge. They lack the mental models to recognize architectural catastrophe. You can prompt your way past syntax. You cannot prompt your way past a fundamental lack of engineering intuition. The hype told us that learning to code is dead because language is all you need. The data just proved the opposite. To truly master the vibe, you still need to understand how the machine thinks.
312
1,317
385
7,150
806,651
things AI has solved: • writing code things AI has not solved: • knowing what code to write • engineering • product design • requirements gathering • the requirements changing after gathering • stakeholders • the stakeholder who "just had a thought" • estimation • the meeting about the estimate • the meeting about the meeting • naming things • cache invalidation • off-by-one errors • the staging environment • whatever DNS is doing right now • Jira • the ticket that's been "in progress" since 2023 • code review comments that are actually about ego • the intern's force push • timezones • dates in general • the printer
81
198
28
1,788
79,193
Seeing internships become fewer at many tech companies tells me a bunch of stuff: 1. No plans to convert a year out. You run internships when you know a year later you’ll do new grad hiring 2. Not enough capacity for mentoring. Interns need at least one mentor - which takes time away from an eng doing whatever they do. On larger teams there’s more “slack” time. On small teams, rarely 3. Team dynamics less important. Interns really boost morale on teams - found this underrated, always. 4. Career growth not important. Mentoring interns grows that mentorship skill that is (used to be?) important for seniors. Food for thought on all
77
74
17
1,896
136,707
Kingdom Hearts IV launches late 2027 - wishlist now! Nintendo Switch 2: sqex.link/4s3v PS5: sqex.link/opzy Xbox: sqex.link/8b2b Steam: sqex.link/zzdw EGS: sqex.link/8wli
251
4,458
419
26,839
836,940
erekhron retweeted
We're making Grok Bot more widely available. All SuperGrok Plus, Cursor Pro+, and Cursor Teams subscribers now have access. We're also offering a free trial with limited usage for all other users.
1,165
1,041
747
13,456
26,185,986
youtu.be/78BTru6W7wI Absolute game changer. This is crazy. Soundpaint is amazing
44
erekhron retweeted
Made a little something LOL
18
942
8
6,044
75,844
The moment Kingdom Hearts 4 and the Disney+ series were revealed! #D23Expo
8
630
43
4,062
112,018
The hype is insane 😭
D23 is RIGHT AROUND THE CORNER! THE KINGDOM HEARTS HYPE IS HIGH! KINGDOM HEARTS 4 IS UPON AND THE LOST MASTER ARC IS SET TO BEGIN! Here is one of my proudest Kingdom Hearts Video edits to help set the mood! Enjoy #KingdomHearts #KingdomHearts4
50
45 mins of new unreleased music + other cool tunes paired with visuals all done by me Premieres tomorrow 9:00am pst
3
2
1
22
1,227
Learn state machines and the actor model. Thank me later.
Are we still talking loops or did we shift to graphs yet?
75
65
17
1,577
199,293
Can't wait for Grok 4.5 to come to Europe. I love what I'm seeing so far and next to composer 2.5 it's gonna be a blast
50
erekhron retweeted
Replying to @coreyhainesco
... I'm 90% conviced these insane cost numbers are made up numbers. Or heavily bloated by investment ratios etc. If it is really that expensive what are doing man. Can't be justified. Then the only research everyone shall invest is to get the costs down.
1
97
erekhron retweeted
New CursorBench results just dropped and Grok 4.5 is the story. #3 overall at 66.7%. Right behind Fable 5 Max at 70.5%. Now look at the cost column. Fable 5 Max: $17.32 per task Grok 4.5 High: $1.51 per task That is Fable level performance at roughly 1/10th the cost. And it beats Fable 5 High and Opus 4.8 Max outright. The intelligence war just became a price war.
179
368
147
4,627
380,298
The whole promise of Claude Fable is I think falsy. Use the biggest best model with a limited time price! Just here, just now! Then we will make you pay even harder for it once you are addicted. This feels just a marketing scheme. I'm not saying it'as bad thought behind is shady
2
36
A full user story that would have took 5-6 months with a full stack team from design to finish in the old way.. Now it is 3-5 days plus testing.. And we can scale it up with AI native SDLC with more engineers. Go to market now is within 2 weeks and not years.
11
still switched back to Composer 2.5 & GLM 5.2. Were more then enough. I hear that Cursor & xAI plan to release there latest versions which are even further improved.. It will be enough. Without Fable and our own curated WeBrain AI harness I also arrived at a 10-30x productivity.
27