Joel Dean
@sirjoeldean
AI Founder | Crafting @joyfulhq an AI-powered mental wellness app | ex @tumblr @wordpressdotcom
The new ChatGPT Advanced Voice Mode can talk in Patois really well 🇯🇲🇯🇲🇯🇲🇯🇲
Claude Code Max is definitely worth it.
whatever is keeping you awake past your bedtime is not worth it
One of the biggest things I monitor with algorithmic feeds is that every feed is tuned to kind of match your reality. So, if you really think about it, your feed has you in this like reality distortion field. So, yeah, it's not really real. It's always good to step outside of…
Coders will have setups like this and not be running Claude Code, Cursor, Windsurf and Gemini CLI all at once.
Men will have entire setups like this just to underperform the S&P 500
Thats the bet I made on mobile 12 years ago.I decided I was going to go all in and learn everything about Android. Attended conferences in Boston, NYC, London etc to meet other Android Engineers, then went on a React Native team. Now I have the experience to build anything.
Pick one thing. Be dangerously good at it. Repeat it everyday like your future depends on it. In a matter of a few months, you'll be unstoppable, my friend.
Agreed. I think one of the best things it does is give you the freedom to work without thinking about api limits or costs. It is also so agentic so you can plan and have it work endlessly knowing that Opus is great at coming up with high quality output once the plan was sound.
If you're not providing every engineer on your team with a Claude Code Max 20x plan, you're falling behind. Teams that use it effectively will significantly outpace those who don't.
Exactly, I got Claude Code Max to use Opus daily and its been a great experience. High quality output once I prompt it well and does a good job of managing context. The best model for code right now based on what I am trying to get done.
The only LLM I use these days is Claude Opus 4. OpenAI doesn't have anything interesting. Gemini is too corporate. Grok feels unserious. The rest are all behind. Opus 4 is by far the best for writing code and has been very good for almost anything else I've thrown at it.…
I’ve been using React Native for ~5 months now, after 4 years of professional development with Flutter and 2 with Swift. Still amazed how often I find that @expo and the community have already thought about so much things to improve quality of life. Feels like magic,…
React Native doesn't get enough praise for what it did for mobile. Spent 12 years doing Android natively at scale and I doubt I will need to touch that kinda code any time soon due to the advantages of using Expo/React Native. May contribute some fixes to Expo Android issues
React Native doesn't get enough praise for what it did for mobile. Spent 12 years doing Android natively at scale and I doubt I will need to touch that kinda code any time soon due to the advantages of using Expo/React Native. May contribute some fixes to Expo Android issues
My favourite logic to write in apps, is scroll logic. It can be one of the most complex things to get working but when it does, the user experience is great. I got a lot of experience doing this while working on page/post editors powering Tumblr & WordPress.
LOL most of these are AI generated
LinkedIn is now processing 11,000 job application submissions per minute—a 45% surge from last year, per NYT
It is better at writing code and the Max plan fits better with my budget. I like to feel like I have unlimited mode to work for a whole month at one set price.
how did claude code get so popular? didn't realize so many devs like using the terminal
Madddddd 🔥🔥🔥🔥🔥
9.86s!!🤯🔥 Oblique Seville 🇯🇲 beats Noah Lyles 🇺🇸 in his Diamond League debut in London, clocking a time of 9.86s (-0.6) to win! Lyles was 2nd in 10.00s, while Zharnel Hughes 🇬🇧 finished 3rd in 10.02s. 4. Ackeem Blake 🇯🇲 10.04 5. Akani Simbine 🇿🇦 10.11 6. Jeremiah Azu 🇬🇧…
Not surprised. Now they will use AI recruiting tools to do filtering to get back to baseline.
Employers are buried in AI-generated resumes, per NYT.
Psychosis prevention in LLMs = evals built on core psychology principles and real time monitoring
I don't feel like being in the gym today but here I am cuz only consistency matters with this stuff.
AI loves Patois 🇯🇲🇯🇲🇯🇲 It has been quite fun building an advanced voice agent with Patois & English functionality.
