←
Addy Crezee | thehype. | /function14,610 posts
Addy Crezee | thehype. | /function1 avatar

Addy Crezee | thehype. | /function1

@AddyCrezee

Can't stop building. @function1_conf @thehypedotnews

◷ Joined March 2015▧ 161 Photos and videos
Addy Crezee | thehype. | /function1 avatar
Addy Crezee | thehype. | /function1@AddyCrezee·Mar 10

Building the place Where the AI World Gathers.

A year ago, /function1 was a one-day curated gathering of 600 people. In late 2025, it became a two-day conference with three stages, 150+ speakers, and 10,000+ attendees from 141 countries - that’s 72% of the whole world! In just 11 months, we have grown 16x in scale, becoming one of the most rapidly growing tech conferences in the world. Today, we are more than just an event: a full-fledged ecosystem where the AI world gathers to communicate, exchange ideas, and evolve. The growth happened fast. The impact is real. And this is only the start.

Post attachment
thehype. avatar
⟳ Addy Crezee | thehype. | /function1 retweeted
thehype.@thehypedotnews·2h

Gemini 4 Neon vs Fable 5.2 at LEGO Ninjago 🥷 Fable 5.2 rebuilt it in voxels for $11.19 with every ninja filling the frame in thick fire and ice while Gemini 4 Neon spent $12.93 and left them small on an empty stage More AI model tests -> thehype.news

0:18

Introducing Gemini 4 Neon! Pushing the frontier across domains and punching above its model class in a meaningful way. To me, Neon feels like the right balance of intelligence, speed, and usability. Available now across @GoogleAIStudio, the API, and more!

Post attachment
Addy Crezee | thehype. | /function1 avatar
Addy Crezee | thehype. | /function1@AddyCrezee·Nov 17

these models are making it even funnier than originals lol

GPT-6.2 Astra vs Fable 5.2 at Adventure Time ⚔️ Astra spent $21.03 on Finn and Jake's giant fight in voxels and kept the giant pink while Fable 5.2 built a denser forest for $13.85 but turned him orange More AI model tests -> thehype.news

0:20
thehype. avatar
⟳ Addy Crezee | thehype. | /function1 retweeted
thehype.@thehypedotnews·Nov 17

Seedance 2.5 Flash vs Seedance 2.5 at Spider-Man🕷️ Flash made the street-cam to chest-cam swing for $3.33 on Higgsfield but lost the hero in its traffic-cam shot, while Seedance 2.5 kept him in frame for $4.67 More AI model tests -> thehype.news

0:16

More ideas to try. Less time waiting. ⚡ Introducing Dreamina Seedance 2.5 Flash: results close to flagship Seedance 2.5, with nearly twice the generation speed. For 720p text-to-video and image-to-video generation, pricing is USD 0.152 per second, approximately 34% lower than Seedance 2.5. The API will launch on BytePlus’s ModelArk platform on Sunday, November 15. Watch the video below.

0:46
thehype. avatar
⟳ Addy Crezee | thehype. | /function1 retweeted
thehype.@thehypedotnews·Nov 16

99% of people don’t know how to properly research with AI Bright Data helps avoid restrictions, so your agent gets to crawl the data without interruption Unlock your access with -> brightdata.com/thehype

0:40
thehype. avatar
⟳ Addy Crezee | thehype. | /function1 retweeted
thehype.@thehypedotnews·Nov 16

Qwen4-Max cost MORE than Opus 5.5 and still LOST at Family Guy Both remade Peter's knee meme in voxels with the same prompt and one fix round, and Qwen4-Max spent $9.22 in api costs to Opus's $7.74 as most of its input missed the cache More AI tests -> thehype.news

0:13

📢 Meet Qwen4-Max, your new AI teammate for work, creativity, and discovery! Qwen4-Max is built on a new MoE architecture with native vision and a 1M-token context window. Highlights: - Outperforms Qwen3.8-Max with up to ~50% fewer tokens, while average cost per task drops over 60% in our internal evaluation. - Advances real-world productivity through greater proficiency with professional tools, stronger long-horizon execution, and higher-quality deliverables. - Handles real tasks, including 3D modeling, game development, manufacturing, finance, law, healthcare, and frontier scientific discovery. - $1.5/M tokens for input, $4.5/M tokens for output, and $0.075/M tokens for implicit caching — all lower than Qwen3.8-Max, with implicit caching down by 70%. The API is live on QwenCloud. You can also try Qwen4-Max on Qwen Studio, QwenWork and Qoder. Start building today! 🚀 - Blog: qwen.ai/blog?id=qwen4-max - API: qwencloud.com/models/qwen4-m… - Token Plan: qwencloud.com/pricing/token-… More to come! The open weights of smaller Qwen4 models will be released. Stay tuned. 👀

Post attachment
thehype. avatar
⟳ Addy Crezee | thehype. | /function1 retweeted
thehype.@thehypedotnews·Nov 12

Claude Sonnet 5.5 4x cheaper than Gemini 4 Hydrogen at Megamind 🧠 Hydrogen spent $36.07 on Megamind's entrance in voxels and its Titan melts into a red blob, while Sonnet 5.5 built a cleaner scene for $8.64 Be like Megamind with Bright Data -> brightdata.com/thehype

0:21

Meet Gemini 4 Hydrogen: built for rapid coding, multimodal synthesis, and everyday workflows. It finishes tasks 2.5x faster than 3.8 Flash and surpasses Argon on the @ArtificialAnlys Intelligence Index.

Post attachment
Derya Unutmaz, MD avatar
⟳ Addy Crezee | thehype. | /function1 retweeted
Derya Unutmaz, MD@DeryaTR_·Nov 11

I asked Fable 5.2 for a film about the life of a T cell, immune cells I study. It wrote the script, fact-checked the immunology, built every cell in Blender, rendered 12,480 frames & added the narration & score! I am stunned by the beauty of this 8 minutes 40 seconds video!🥹😍

8:40
thehype. avatar
⟳ Addy Crezee | thehype. | /function1 retweeted
thehype.@thehypedotnews·Nov 10

What if we broke the speed of light at the Large Hadron Collider? We pushed the beam past 100% – the pipe tears, then Geneva caves in. Built in three.js by Fable 5.2 More AI model tests -> thehype.news

0:49
thehype. avatar
⟳ Addy Crezee | thehype. | /function1 retweeted
thehype.@thehypedotnews·Nov 10

Gemini 4 Carbon LOST to Fable 5.2 at Peppa Pig 🐷 Both rebuilt the Peppa Pig opening titles as a 3D voxel scene from the same clip – Carbon cost $10.10 in API tokens and Fable $19.60, yet only Fable's pigs look like the show More AI model tests -> thehype.news

0:15
Google@GoogleNov 10

Gemini 4 Carbon is now available. It’s our strongest model, built to dig deeper, handle long-horizon tasks, and multi-step reasoning. If Argon is great for complex tasks, Carbon is the model you bring in for your hardest problems.

0:05
thehype. avatar
⟳ Addy Crezee | thehype. | /function1 retweeted
thehype.@thehypedotnews·Nov 8

rumor: after alibaba teased qwen 4 on nov 6, the first benchmarks of the new model started to surface the developer who posted the results says alibaba gave him early qwen 4 checkpoints. he ran them on gamefeel bench, a game-dev benchmark focused on coding and creativity worth flagging: there are no other credible benchmarks yet. it is one developer's benchmark, opus 5.5 checks part of the answers, and the claude prices are subscription-based while qwen runs on api. alibaba has not commented, and its only official word on qwen 4 so far is the nov 6 teaser well, we can't wait for the new qwen to drop – curious where it lands on the @ArtificialAnlys intelligence index

Post attachmentPost attachment
thehype. avatar
⟳ Addy Crezee | thehype. | /function1 retweeted
thehype.@thehypedotnews·Nov 5

mistral, europe's most valuable ai startup, added managed agents and apps to studio – a platform where companies build agents on mistral large 4 (le chonk) and run them on their own servers @MistralAI says a workflow it built there saves its product ops team about 35 hours a week. the model tags bugs and feature requests, then routing rules send each one to the right team agents get skills, code sandboxes, subagents and hooks. they run on the runtime that already handles millions of mistral vibe sessions a week both the models and studio can be deployed on-prem or in your own vpc. openai and anthropic don't offer that – their models only run in the cloud. each run leaves opentelemetry traces, and evals check prompt or model changes against your own datasets

Post attachment

🆕 Le Chonk brings the brains. Mistral Studio gets it working. Introducing managed agents and apps, plus new capabilities for workflows, enterprise connections, observability and evaluations in Studio. Build faster. Deploy where you choose. Keep control. See what’s possible. 🧵

0:30
thehype. avatar
⟳ Addy Crezee | thehype. | /function1 retweeted
thehype.@thehypedotnews·Nov 4

openai is buying pierre computer, a dev-tools startup whose tech already runs inside codex - axios calls it a direct challenge to microsoft's github pierre's main product is code storage, a git host that ai agents work with through an api. agents create repos and push code there on their own, with no rate limits lovable and bolt, which build apps from a prompt, keep their generated code there. so does the coding agent amp. the ai assistant poke stores its agent memory there too openai will fold code storage into codex, so the code codex writes can be stored, reviewed and shipped in one place. terms weren't disclosed. pierre's open-source tools, diffs and trees, stay open

Post attachment
thehype. avatar
⟳ Addy Crezee | thehype. | /function1 retweeted
thehype.@thehypedotnews·Oct 31

top 5 github repos by weekly star growth: the skills meta is back – the top 3 are all agent skills, 1,515 stars apart 1. cloudflare/security-audit-skill by @cloudflare – +12,838/week, 54.8k total. agents audit code, a fresh agent tries to disprove every bug 2. nutlope/hallmark by @nutlope – +11,766/week, 43.5k total. anti-ai-slop design skill for claude code, cursor and codex 3. pbakaus/impeccable by @pbakaus – +11,323/week, 97.8k total. 24 design commands and a 62-rule slop detector 4. affaan-m/ecc by @affaanmustafa – +6,506/week, 293.7k total. skills, memory and security hooks for coding agents 5. tauricresearch/tradingagents by @TauricResearch – +5,490/week, 118.9k total. multi-agent llm trading framework two of the top 3 fight ai slop in generated interfaces. this week's #1 pulled 8x fewer stars than rea did last week follow @thehypedotnews for weekly github trending and 24/7 ai radio

Post attachment
0:11
Addy Crezee | thehype. | /function1 avatar
Addy Crezee | thehype. | /function1@AddyCrezee·Oct 30

I’ll pay for terra first, and improve with opus after

gpt-6.1 terra rebuilt the fury road war-party chase 8x cheaper than opus 5.5 5 shots in three.js, every car and explosion in code opus 5.5 – 59m, $21.90 gpt-6.1 terra – 1h 6m, $2.70 would you pay 8x more for the opus cut?

0:35