Dylan Halliu (@Dylan_Halliu) header
Dylan Halliu (@Dylan_Halliu) profile picture

Building AI products @Microsoft, Co-Host of @FeatureCrewPod All thoughts are own!

NYC
Joined June 2019
Photo by Dylan Halliu (@Dylan_Halliu)Photo by Dylan Halliu (@Dylan_Halliu)Photo by Dylan Halliu (@Dylan_Halliu)Photo by Dylan Halliu (@Dylan_Halliu)Photo by Dylan Halliu (@Dylan_Halliu)Photo by Dylan Halliu (@Dylan_Halliu)Photo by Dylan Halliu (@Dylan_Halliu)Photo by Dylan Halliu (@Dylan_Halliu)Photo by Dylan Halliu (@Dylan_Halliu)Photo by Dylan Halliu (@Dylan_Halliu)
Welcome to our live show! nitter.cf/i/broadcasts/1YqKDNeZn…
2
3
81
We're going live at 5:30est/2:30pst today to do our AI predictions for 2026! This will be the first of weekly livestreams on The Feature Crew See you in a few hours here on X and YouTube!
2
1
4
387
Hearing rumors that the new o3 in @ChatGPTapp is a preview of gpt-5. It is significantly better at our planet test compared to o3 at launch. Results below look much more like @petergostev 's results from the lmarena mystery models EXCITING
10
7
5
117
28,422
OpenAI is cooking
OpenAI are testing a new model on the Web Dev Arena @lmarena_ai under the name 'Anonymous Chatbot 0717'. I can't believe I'm gonna say this, but it is genuinely at a completely different level of front end coding - far better than Sonnet, o3, Gemini 2.5 Pro, or Grok 4. To test it, I ran a great prompt borrowed from the amazing @FeatureCrewPod YouTube channel, asking models to create a procedurally generated planet with Three.js. Take a look yourselves, but I'm pretty astonished by how big the jump is. I have featured the new model twice just because its implementations have been so interesting. Of course, this is only one test, but OpenAI models have always been a bit 'meh' at front-end work, and they seem to have finally overtaken everyone else on that front. We'll see when it comes out. Credit to @chetaslua for discovering the model
1
3
12
1,340
Hey! Our little AI podcast is at 2.3k subs! Give it a chance if ya like seeing models get put through their paces.
1
1
2
76
Super interesting analysis It looks like images may confuse o3 while reasoning over certain tasks. Thanks for the investigation Jacob!
Replying to @FeatureCrewPod
First, I love the pod. You are top on my goto list for each model release. Second, there was a comment on YouTube suggesting to only use the image, not the ascii, so, I gave it a go. It was an agonizing 9 minutes and 53 seconds, but it figured out the 20x20. (ChatGPT doesn’t share images, so, i attached them to this post!) chatgpt.com/share/68006b23-9…
1
2
3
211
4.1 did have a good showing for coding as a one shot- probably the best one shot coding performance we’ve seen
🚀 Exciting news! OpenAI has launched GPT-4.1, and we're diving deep into its features and capabilities. Check out our latest video to catch the vibe! 🎥✨ #GPT41 #AI #MachineLearning youtube.com/watch?v=k3rSDHe3… youtube.com/watch?v=k3rSDHe3…
1
65
/ 3月1日(土)から フォロー&リポストキャンペーン🌸✨ \ #アサヒビールミュージアム では、春限定のスペシャルな #泡アート をご用意しています🍺 ① @asahibeermuseum をフォロー✨ ② このポストをリポスト ③ 画面をクルーに見せる👀 春を感じる特別な泡アートをぜひお楽しみください🍺
8
531
1
154
25,372
Super interesting results for our first look at - Jakob/Chris had some great predictions on the agentic-ness of the new Sonnet 3.7 model!
🚀 Excited to share our first reactions to the new Claude Sonnet 3.7 Model! We dive into coding and agentic reasoning. Check it out and stay tuned for head-to-head comparisons! 🎥✨ #AI #Claude #MachineLearning #FeatureCrew youtube.com/watch?v=QtAODwyR… youtube.com/watch?v=QtAODwyR…
1
74
🚀 Check out our latest video where we test the new Deep Research feature in ChatGPT for high-level archaeology, product development, and algorithms! How does it compare to other research agents? 🤔🔍 Watch now! 👉 youtube.com/watch?v=tZOtWJgK… #DeepResearch #… youtube.com/watch?v=tZOtWJgK…
1
1
3
314
We got access to Deep Research as well - we'll get a video out soon!
50
Crazy times, a trillion $ datacenter isn’t that insane to talk about anymore (but still 4x NZ GDP…)
Announcing The Stargate Project The Stargate Project is a new company which intends to invest $500 billion over the next four years building new AI infrastructure for OpenAI in the United States. We will begin deploying $100 billion immediately. This infrastructure will secure American leadership in AI, create hundreds of thousands of American jobs, and generate massive economic benefit for the entire world. This project will not only support the re-industrialization of the United States but also provide a strategic capability to protect the national security of America and its allies. The initial equity funders in Stargate are SoftBank, OpenAI, Oracle, and MGX. SoftBank and OpenAI are the lead partners for Stargate, with SoftBank having financial responsibility and OpenAI having operational responsibility. Masayoshi Son will be the chairman. Arm, Microsoft, NVIDIA, Oracle, and OpenAI are the key initial technology partners. The buildout is currently underway, starting in Texas, and we are evaluating potential sites across the country for more campuses as we finalize definitive agreements. As part of Stargate, Oracle, NVIDIA, and OpenAI will closely collaborate to build and operate this computing system. This builds on a deep collaboration between OpenAI and NVIDIA going back to 2016 and a newer partnership between OpenAI and Oracle. This also builds on the existing OpenAI partnership with Microsoft. OpenAI will continue to increase its consumption of Azure as OpenAI continues its work with Microsoft with this additional compute to train leading models and deliver great products and services. All of us look forward to continuing to build and develop AI—and in particular AGI—for the benefit of all of humanity. We believe that this new step is critical on the path, and will enable creative people to figure out how to use AI to elevate humanity.
1
1
36
2025 will be a wild year, from below - “We are now confident we know how to build AGI as we have traditionally understood it. We believe that, in 2025, we may see the first AI agents “join the workforce” and materially change the output of companies. We continue to believe that iteratively putting great tools in the hands of people leads to great, broadly-distributed outcomes.”
1
84
The sudden vigor with which people debate AGI is an indicator of how undeniably close we are
1
4
118
Still don’t know how to process that we have glimpses of AGI - its hard to imagine how much the world will change, especially when most people outside of the AI space have no idea of the implications. What’s everyone else’s take?
23
O3 announcement is incredibly exciting - the performance on ARC is already human level, can’t believe there’s no wall - Congrats to the OpenAI team!
2
1
130
🚀 Dive into OpenAI’s Sora with The Feature Crew! 🎥 Discover the strengths, weaknesses, and creative potential of this amazing video generation technology. 🌟 Check out our latest video now! #AI #VideoCreation #OpenAI 👉 youtube.com/watch?v=m6Epj2ia… youtube.com/watch?v=m6Epj2ia…
1
1
77
If you don't want to pay $200 for unlimited Sora send me your requests and I'll generate. Results are mind blowing
8
1
2
5
3,521
Sora (on plus subscription) seems to not allow uploaded media with people by default. Seems like a good broad RAI mitigation to start - avoids the deep fake generator can of worms
120
Sora storyboard is insanely cool - great UX, will play around with Sora, but even alone this is a massive step up.
1
1
1
110
Dang - already hitting capacity, totally understandable
31