Jerry Beers

@Dimillian Help me write metal shaders

As an iOS dev, what at the iOS dev specific tasks you want Codex to help you better at?

@Rengle820 I’d love to see a collaboration with hackmotion, so I can do my drills in Golf+ when I don’t have room for a full club swing.

Traditional sim long game plus our mixed reality sim short game. Put on or take off the headset whenever you want, totally up to you!

Ever since I used Vision Pro for the first time, I instantly knew that I wanted to develop an app for this beautiful platform! Introducing AuraTap - a place to meet & connect with fellow Vision Pro users using one of the most incredible feature - Personas! 🗓️ Available Friday

You meet someone new on Apple Vision Pro. Face to face. 2 minutes. That’s the idea behind AuraTap, a new app. If there’s a mutual vibe, you keep going. If not, you’re on to the next. I love how quickly you can meet people, make connections, and explore each other’s profiles. It also features one of the most beautiful UX designs I’ve tried on visionOS. Everything feels incredibly polished. The app launches this Friday, March 27. They’re hosting an in app launch party this Friday at 1PM PDT. Hope you can make it.

@JonathanBCrabb How’s it going?

As God is my witness, I am going to get openclaw working this weekend. I will get past the gateway not being able to start and I will get a response from bot on telegram so I can start taking over the world.

@JonathanBCrabb Yeah, I think a lot of things that are amazing, like the virtual environments, wouldn’t be as impressive without the quality of the displays. Did you watch any of the 180° video?

@JerryBeers_ How high fidelity it looked. The super crisp icon / imagery being shown right next to the humans standing next to me was wild. I think the immersive stuff was fascinating but needed more time to get used to the format.

I tried out the Apple Vision Pro last night for the first time. I expected it to be good, but was blown away at some of the clarity and features.

Wow, this is so cool

After a ton of work, it's so great to see these reviews 🤯 Looking for something cool to experience on Apple Vision Pro? Take a break in the Winter Lights environment – it's magical! ❄️✨ Happy holiday, X!

Anybody know of a good ranked list of must-have apps (or even nice to have) for the Vision Pro? @justinryanio

NEW Vibescape for Apple Vision Pro is now available! Winter Lights is our most expansive immersive environment yet. Go chill as snow falls gently, stars twinkle, and a million lights softly glow on a forest of thousands of trees ✨

If this were a film at a flight museum, they'd be charging $20 per viewing, and there would be a line to see it. Honestly so much of Apple's immersive films are so good, but as an aviation enthusiast, this one tops them all. So. Good. Go get a Vision Pro demo and ask to see this film.

@kisalay_Cool95 @GeminiApp I’ve tried this use case with no success. What prompt did you use?

@GeminiApp I tried the collage method with Gemini last week—combined old family photos into one frame. The result felt like a time machine. One prompt, and suddenly generations were side by side in a single image. Absolutely magical. 🌌

Combine your photos in a whole new way with Gemini. Upload multiple images to Gemini, then tell it how to combine them with a single prompt. Here are some of the ways we’ve seen people use the "collage method” with #nanobanana in the Gemini app 🧵

@JonathanBCrabb It’s a great season

Finally getting into season 5 of The Chosen episode 1. This show continues to be a gift.

Playing Balatro is like solving an intricate puzzle, but in the middle you might have to start over with completely different pieces.

“programming languages don’t matter anymore”. THIS. That’s what I’m going to build my upcoming conference talk about. You are no longer an iOS/Web/Backend dev. You are a builder.

I can see it now, Spectrum offers 125 peta bytes per second download speeds, but still 5 Mbps upload speeds. livescience.com/technology/com…

@JonathanBCrabb I think it would be a fascinating challenge

I had a dream I won a $238M lottery. It was a nightmare.

🚨 Big news: We’re headed to AWE USA 2025! 🚨 XelerateVR is joining the world’s leading XR innovators in Los Angeles to showcase how natural movement will redefine what’s possible in VR. If you're attending, come try our omnidirectional treadmill live – and feel what true immersion really means. This isn’t just another hardware demo. It’s a glimpse into the future of gaming, training, and everything in between. Get a first impression in the epic trailer below. 📍 Booth 334, right next to Meta 📅 10.06 –12.06 🔗 xeleratevr.com Let’s connect – we’re open to partnerships, press, and potential investors. Drop us a message if you’ll be there – we can also give away free tickets! #AWE2025 #XR #VR #ImmersiveTech #XelerateVR #HardwareStartup #FutureOfVR

Put together a pretty comprehensive Midjourney parameter guide you can use as reference when prompting. This is where a lot of the MJ magic lives, and how you can start to really tune aesthetics and styles to your exact preferences: ### Basic Parameters ### Aspect Ratio (--ar) Format: Whole number pairs (e.g., `--ar 16:9`, `--ar 1:1`) Default: --ar 1:1 (square) Function: Controls the w:h ratio of generated images Common Values: 16:9 (widescreen), 9:16 (vertical), 3:2 (landscape), 2:3 (portrait), 4:5 (Instagram) Stylize (--s or --stylize) Range: --s 0 to --s 1000 Default: --s 100 Function: Controls balance between prompt adherence and creative interpretation. High values inject more of the underlying model aesthetic when used without personalization. With personalization on (--p), high values inject more of the personalization aesthetic into generations. Lower Values (0-100): Stricter prompt interpretation, less creative interpretation Higher Values (100+): More creative interpretation, less literal adherence to prompt Chaos (--c or --chaos) Range: --c 0 to --c 100 Default: --c 0 Function: Controls level of variety between images in the same generation Lower Values (0-20): Similar images across the grid Higher Values (40+): Dramatic differences Recommended Range: 20-40 for balanced variety without breaking style and consistency Raw (--raw) Values: Parameter flag (no numerical value) Effect: Favors more literal interpretation of prompts with reduced creative liberties. Also leans towards more photorealistic imagery with less stylization when medium/style is not mentioned in the initial prompt. Experimental (--exp) Range: --exp 0 to --exp 100 Default: --exp 0 Function: Adds detail and dynamic elements to images Optimal Range: 5-25 for balanced results High Values (40+): May drift significantly from prompt Sweet Spot: --exp 5-20. Higher --exp values are best combined with higher --stylize values Weird (--w or --weird) Range: --w 0 to --w 3000 Function: Introduces unusual elements and disrupts normal aesthetics Note: Sometimes breaks style references and personalization when used in same prompt Quality (--q or --quality) Options: --q 0.5, --q 1, --q 2, --q 4 Function: Generations use more (or less) compute for each job in attempt to Usage: Higher quality settings take longer to generate and cost more, but may produce better results Seed (--seed) Range: Any number (practically unlimited) Function: Locks the initial noise pattern for repeatable results with same prompt Usage: Useful for testing parameter changes while keeping initial conditions identical Note: Only works with initial generation; variations create new seeds Tile (--tile) Function: Creates seamless patterns for tiling textures Usage: Useful for creating repeating backgrounds and textures No Parameter (--no) Usage: Excludes specific elements from generation Format: `--no [element to exclude]` Example: `--no color, object, texture` Note: Functions inconsistently in V7 Repeat Parameter (--r or --repeat) Range: --r 2 to --r 40 Default: Not specified (runs once if not included) Function: Runs the same prompt multiple times Usage: Useful for quickly generating multiple sets of images from the same prompt Example: --repeat 3 would run the same prompt 3 times Application: Often used with style reference exploration (e.g., --sref random --r 10) Note: the total number of prompts you can run at once depends on your plan. Basic plan up to --r 4, standard plan up to --r 12, pro and mega up to --r 40 ### Reference Systems ### Style Reference (--sref) Format: Style code, image reference, or --sref random Function: Applies a style/aesthetic treatment to your generations based on the code or image used. When using image, it decouples subject from aesthetic style Note: ~4.6 billion style reference codes. Use '--sref random' to have one randomly assigned to your prompt. Usage: Can blend multiple styles using space separation Example: `--sref code1 code2` Blending Syntax: Use double colon (::) followed by number to control relative weights of each reference Example: `--sref code1::3 code2::1` (3 parts of code1, 1 part of code2) Style Weight (--sw) Range: --sw 1 to --sw 1000 Default: --sw 100 Function: Controls intensity of style references Lower Values (25-50): Subtle style application Higher Values (250-1000): Stronger style influence Note: Only affects style references, not personalization or mood boards Image Weight (--iw) Range: --iw 0.25 to --iw 3 Default: --iw 1 Function: Controls balance between image influence and text influence in image prompts Values: --iw 0.5: Text prompt twice as important as image --iw 1: Equal importance --iw 2-3: Image has 2x or 3x more influence than text Recommended: --iw 0.5 to --iw 1.5 for balanced results in most cases Omni Reference (--oref) Format: Image placed in 'Omni-reference' area of prompt bar Function: References subjects, objects, characters, and more and integrates them into your generations (expanded character reference) Note: Available only in --v 7 Usage: Used with Omni Weight parameter for control Omni Weight (--ow) Range: --ow 1 to --ow 1000 Default: --ow 100 Function: Controls intensity of omni reference Lower Values (25-50): Better for style switching Higher Values (250-400): Better detail preservation of referenced elements from image ### Personalization & Mode Controls ### Personalization Profiles (--p or --profile) Function: Tunes base model to user's aesthetic preferences based on users image pair rankings in personalization tab or moodboard creation Requirements: Minimum 200 ratings to unlock for basic functionality, 1000+ ratings for optimal results Note: Global profile code adjusts with new ratings, but your most recent can always be added using --p. Moodboard codes also follow the --p parameter. Draft Mode (--draft) Function: 2x faster processing at 50% cheaper cost Quality: Slightly different aesthetic from standard mode Usage: Ideal for rapid iteration Note: Can be enhanced to standard quality Conversation Mode Features: - Natural language prompting - Voice command capability - Real-time prompt updates - Support for follow-up modifications - Can generate multiple variations - Limitations: Some parameters need to be typed rather than spoken ### Advanced Techniques ### Permutations Use case: run multiple iterations of prompt in one go Syntax: Curly braces with comma-separated options Example: `{option1, option2, option3}` Limits: - Basic plan: Up to 4 permutations - Standard plan: Up to 10 permutations - Pro plan: Up to 40 permutations Advanced Use: Nested permutations with double parentheses Example: `{man wearing {white tee, black shirt}, woman wearing {red dress, camo jacket}}`