DopamineDevs (@DopamineDevs) header
DopamineDevs (@DopamineDevs) profile picture

AI, coding, and emerging tech. A space built for developers who live to code, learn fast, and connect with fellow devs to shape the future

Joined July 2025
Photo by DopamineDevs (@DopamineDevs)Photo by DopamineDevs (@DopamineDevs)Photo by DopamineDevs (@DopamineDevs)Photo by DopamineDevs (@DopamineDevs)Photo by DopamineDevs (@DopamineDevs)Photo by DopamineDevs (@DopamineDevs)Photo by DopamineDevs (@DopamineDevs)Photo by DopamineDevs (@DopamineDevs)Photo by DopamineDevs (@DopamineDevs)Photo by DopamineDevs (@DopamineDevs)
Welcome to Dopamine Devs. We’re not here to flood your feed. We’re here to filter it. This is where builders, creatives, and curious minds come to swap doomscrolling for signal. Expect quick, clean rundowns on startup and AI moves, tools that actually help you ship faster, and spotlights on devs and creators doing dope things, not because they paid to be here but because they’re worth knowing about. No BS. No fake urgency. Just stuff we’d want in our own group chats. Useful drops, smart recaps, and ideas that make you want to open your laptop. This isn’t about a personal brand or selling courses. It’s about showing up, building together, and staying sharp in a world that’s getting noisy fast. We’re just getting started. If you’re working on something, you’re one of us. More to come soon, stay tuned and hit that follow button!
1
1
3
126
Google × Kaggle just dropped a farming sim where your Al agent competes head-to-head. 720 turns per match. Your bot manages crops, animals, and trades on a dynamic market. $5OK in prizes across 10 spots. Deadline is September 23, 2026. kaggle.com/competitions/kagg…
16
DopamineDevs retweeted
I’m starting to realize these Fable 5 and Claude Code rate limit extensions aren’t just random acts of kindness. Anthropic says Fables uncertain availability is caused by limited compute. Sure I believe that, but at the same time every extension also gives users more time to build it into their workflows. Understandings its depth, and becoming familiar with an output that Claudes weaker models cannot fully replicate. That matters because Anthropic is not operating without alternatives. GLM, Kimi, DeepSeek, and other cheaper models are closing the gap. I have not personally compared all of them so I can’t claim they match fables across the board. But they do not necessarily need to. They only need to become good enough for enough tasks, at a lower cost, to weaken its hold on users. If fable disappeared from the $20/m plan tomorrow, I probably would not abandon Claude. I actually respect their R&D and consider it one of the strongest AI teams in the market. However, I would move part of my workflow elsewhere because no single company offers the best tool for everything. No one glove fits all. This is where the strategy becomes interesting. Anthropic is intentionally giving users enough access to appreciate how capable Fable is. Once people reorganize their work around that quality, removing it creates a much greater willingness to pay extra for the same output. At the same time, Anthropic may have accidentally made itself dependent on Fable because it has become one of the biggest reasons customers remain subscribed. With Anthropic moving toward public markets and AI’s underlying economics becoming more expensive than many expected, these temporary extensions may be about more than access. They may also be about protecting habits, retaining users, and buying time before the business model becomes less generous. Fable may be Anthropic’s greatest moat right now. It may also be the product Anthropic can least afford to remove. Would you continue paying for Claude without Fable, or would that push more of your workflow elsewhere?
We're extending Claude Fable 5 access on all paid plans, as well as keeping Claude Code’s weekly rate limits 50% higher, through July 19.
4
2
8
234
Everyone please watch his video if your overwhelmed in the AI space.
Just broke down the AI tools I actually use and why. It's easy to drown in the news. Give this a read.
1
103
So this is huge… what are you guys gonna do with the extra usage?
We’ve agreed to a partnership with @SpaceX that will substantially increase our compute capacity. This, along with our other recent compute deals, means that we’ve been able to increase our usage limits for Claude Code and the Claude API.
1
1
2
27
Just made a custom lil pet agent from our logo
To create your own pet, install the hatch-pet skill:
4
483
DopamineDevs retweeted
Meet Codex for Students. We're offering college students in the U.S. and Canada $100 in Codex credits. Our goal is to support students to learn by building, breaking, and fixing things. chatgpt.com/codex/students
321
234
143
3,666
884,933
Just let the builders use openclaw with antigravity. Why must you be so strict
To the builders: we heard you. We're welcoming back everyone who recently had their Google Antigravity accounts restricted for use of third-party tools. Moving forward, we’ll have clear steps for users to restore their account if it’s restricted. To maintain the integrity of Antigravity and ensure a great user experience for everyone, using third-party tools with your Antigravity login remains against our terms. We love seeing innovation and boundary-pushing in this community, and the Antigravity team is hyper-focused on building what you need to accelerate product development. Can’t wait to see what you build!
1
53
Your 2026 hackathon lineup Stop scrolling. Start building. dev.events/hackathons for more
1
1
58
Interested to see how it’ll handle 1m context window. Will it drown or have they optimized the model?
Introducing Claude Opus 4.6. Our smartest model got an upgrade. Opus 4.6 plans more carefully, sustains agentic tasks for longer, operates reliably in massive codebases, and catches its own mistakes. It’s also our first Opus-class model with 1M token context in beta.
2
2
142
DopamineDevs retweeted
Introducing Claude Opus 4.6. Our smartest model got an upgrade. Opus 4.6 plans more carefully, sustains agentic tasks for longer, operates reliably in massive codebases, and catches its own mistakes. It’s also our first Opus-class model with 1M token context in beta.
1,656
4,613
2,896
38,611
10,628,858
Google × Kaggle ran a 5-day intensive course where the goal was simple: build agents that actually work. Full list of winners + code here: kaggle.com/competitions/agen… #aiagents #agenticai #developers #buildinpublic #kaggle
1
1
27
DopamineDevs retweeted
Software Engineering Expectations for 2026 - The majority of your code should be written by AI now - Cursor/Codex/Claude Code/Gemini/etc - You should try all the tooling and switch between them, as each one gets an edge over the others depending on the release cycle. - You should be using AI to check the code that is written by AI - Have AI write tests - Have AI read logs - Have AI navigate your browser - I don't do this every time because sometimes it's simple enough to check it myself - You should still skim code changes - This can be a lighter skim on internal tools and a heavier read through on customer facing code - Use AI to help you define specs
297
519
71
6,564
536,148
Anthropic gave Claude a credit card + a business. It didn’t fail from “bad answers.” It failed from bad structure.
1
2
11
Dev takeaway: stop using one prompt to do everything. Pattern: Task agent + Oversight gate + Executor. Treat autonomy like production access.
1
1
12
Question: what’s the first permission you revoke from an agent? Pricing? refunds? purchases? supplier emails?
1
3