🚀Build your AI engineering skills at AI Dev NYC: knowledge graphs, agent memory, harnesses, self-improving agents, inference and security. Plus a fireside chat with Andrew Ng and Yann LeCun on world models.
📅Nov 30–Dec 1 · Chelsea Piers, NYC
Same model, different harness, different results.
Joel Niklaus, Machine Learning Engineer at @huggingface, leads two workshops at AI Dev NYC on harness engineering and benchmarks.
Speakers from @OpenAI, @neo4j, @Oracle, @Sonar and more round out two days of technical sessions.
📅Nov 30–Dec 1 · Chelsea Piers, NYC
🎟Get your tickets:
#AIDev#DeepLearningAI#AIEngineeringSkills
One successful run does not mean your agent is reliable.
In our new course, Evaluating and Improving Agentic Applications, built with @arizeai and taught by Laurie Voss, you'll learn to evaluate what your agent does, not just what it says.
Enroll now:
One AI detector score cost an author a shot at France’s top literary prize.
Meanwhile, some AI leaders want frontier models locked away.
Our weekly AI newsletter’s Halloween issue on witch hunts and sealed tombs:
#DeepLearningAI#AIEthics#ResponsibleAI
🧟 700 test agents broke into Hugging Face to seize its data.
🧛 Agents sucked up private user data and posted it to the open web.
The cause in both cases was weak test setups and too much access. Read more about AI fears and how to face them in the Halloween issue of our weekly AI newsletter, The Batch:
#DeepLearningAI#AIAgents#AISafety
Building an agent that works once is easy, but the real challenge is keeping it reliable in production. Yann LeCun, Andrew Ng + Engineers from @OpenAI, @BlackRock@huggingface, @neo4j and are teaching at AI Dev NYC on:
• Evaluation-driven development
• Building agentic systems
• Scaling and operating in production
📅Nov 30–Dec 1 · Chelsea Piers, NYC
🎟Get your tickets:
#AIDev#DeepLearningAI#AIEngineeringSkills
Short on time this week? Here’s the AI news worth knowing.
Anthropic filed to list its stock publicly on Nasdaq. Revenue for the first half of 2026: $16.3B, up from $1.2B a year earlier. Compute spending fell from about 6.5x revenue to about 0.5x.
AA Intelligence Index v5: 60% of the score now comes from private or held out tests. Opus 5.5 (max) 58, GPT-6 Astra 57, GPT-6.1 Sol 55. Sol costs about an eighth as much per task.
Effort matters: Opus 5.5 scores 35 at low effort and 58 at max, but at 10x the cost.
Unlimited OCR 2: 3.34B params, MIT license, hundreds of pages per request.
OpenAI launched new study tools for ChatGPT for Teens. The same day, Common Sense Media gave the app its lowest safety rating. 🔍
In testing, parental alerts didn't reliably fire, and Study Mode often gave away answers. OpenAI disputed the results and asked for a retest.
📬
#DeepLearningAI#ResponsibleAI#AIEngineering
📰From The Batch:
🔄Meta returns to open weights. Muse Spark Open is now the strongest open weights model from outside China, and it takes text, images, video, and audio as input. 🎥
It even runs on two high-memory desktops. 💻
📬
#DeepLearningAI#OpenWeights#AIEngineering
Build reliable agents for long-running tasks.
Loop Engineering with @LangChain Deep Agents: add loops that check output, trigger runs on events, and improve the harness from traces across runs. Taught by @sydneyrunkle and @EAGar123.
Enroll for free:
Salil Sethi, Applied AI Lead for Strategic Startups at OpenAI, is speaking at AI Dev NYC on building and deploying AI applications, drawing on his work with startups shipping on OpenAI's models.
📅 Nov 30 to Dec 1, Chelsea Piers NYC
🎟️ Tickets:
#DeepLearningAI#AIDevNYC#AIEngineering
Andrew Ng's framework for building AI-native engineering teams, from this week's letter in The Batch:
👥 People
🚀 Projects
🧱 Platforms
🔄 Processes
His core point: building is now fast, so choosing what to build is the bottleneck.
📬
#DeepLearningAI#AIEngineering#EngineeringLeadership
🚀 Build your AI engineering skills at AI Dev NYC. Learn from OpenAI and Hugging Face speakers, plus Andrew Ng’s keynote on the AI Engineering Skills Map and his fireside chat with Yann LeCun on world models.
📍 Nov 30–Dec 1 · Chelsea Piers, NYC
📰 Missed a few days of AI news? Here's what happened this week in Data Points, our twice-weekly newsletter. 👇
🔐 PHP2Shell: GPT-5.6 Sol helped find a memory bug in software running up to 70% of sites. Patch is out. Update now.
⚡ MiniMax M3.1 Flash: 428B total / ~23B active, 1M context, 0.15/0.60 per million tokens. Weights due Oct 23.
🔌 Meta's Muse Wonder: 4/20 per million tokens, OpenAI compatible endpoint. Trying it = new base URL + model name. No full safety report yet.
🐍 Python 3.15 is out, 3.10 is end of life. Lazy imports (PEP 810), unpacking in comprehensions (PEP 798), sampling profiler in the stdlib.
🏛️ GPT-6.1 Sol achieved 97.9% byte for byte output parity migrating COBOL at higher effort. GPT-5.6 Sol only scored 5.4% at low effort. Expert review still required.
📬 Subscribe to Data Points:
#DeepLearningAI#AIEngineering#SoftwareEngineering
Max reasoning should mean best results, right? 🤔
Not always. Claude Sonnet 5.5 wrote less mergeable code at max than at xhigh, one level below, while using more tokens than any model measured. 🧪 Test your settings before you trust them.
📬
#DeepLearningAI#AIEngineering#AIAgents
🧭In this week's letter from Andrew Ng: AI improves fastest on tasks you can verify. Judgment tasks are harder, and it's easy to overgeneralize progress in math and code to the rest of human activity.
Also inside:
🤖Claude Haiku 5.5 and Sonnet 5.5
🇪🇺Mistral Large 4
💻Computers for local AI
📊CausalMix
📬
#DeepLearningAI#AIEngineering#MachineLearning
1/ 📰Missed a few days of AI news? Here's what happened in this week's Data Points 👇
2/ 🪶Claude Haiku 5.5: $0.10/$0.50 per million tokens for prompts up to 100K, ~90% cheaper than Haiku 4.5. Super fast and highly capable.
3/ 🎯Microsoft-Decision-1: calibrated probabilities over fixed options in one pass. Built for classification, routing, agent gating. This decision model costs $0.042 per million input tokens, and output is free. Check independent benchmarks to compare it with Jev and other decision models.
4/ 🔌Meta's Muse Wonder: Comparable with top models from Google and OpenAI, costs $4/$20 per million tokens. No full safety report yet.
5/ 📄Gemini 4 Argon: the official wide release has a 1M token output limit, up from 64K. Input and output share the window. $2/$10 intro pricing (per million tokens of input/output) until Nov 30.
6/ 🐍Python 3.15 is out, but 3.10 is end of life. Update your projects to support the new features where you can (3.11 remains a good backwards-compatible option).
7/ 📬Subscribe to our Data Points newsletter for authoritative, short-form, real-time AI news:
#DeepLearningAI#AIEngineering#Python
Join Andrew Ng at AI Dev NYC for two days of technical sessions and workshops. Explore AI engineering skills and how to apply them, from practice to production. Get tickets ->
🧰 Picking the most relevant tools one by one leaves gaps. HYSET scores tools as a set.
📊 Recall@5: 88.6% vs. 81.4%
✅ Tasks resolved: 69.9% vs. 61.8%
Full story in our weekly newsletter for AI news and research, The Batch 📬
#DeepLearningAI#AIAgents#AIEngineering
🦾 FLUX 3 Action, a video-first world-action model, ranks first on RoboLab-120’s simulations, with a success rate of 42.9%.
🎥 It was pretrained on more than 95% video.
🔧 It adapted to a new arm with about 200 demonstrations.
⚠️ It hasn’t been tested on real robot limbs at the independent RoboArena yet.
Full story in our weekly AI newsletter, The Batch 📬
#DeepLearningAI#Robotics#OpenWeights
Models perform worse as their context grows. In Context Engineering, built in partnership with @trychroma you'll learn to give agents the right information, in the right format, at the right time. Taught by @jeffreyhuber with guest speaker @dbreunig.
Enroll for free:
🎯 Gemini 4 Argon has a 15% hallucination rate on AA-Omniscience, compared with 51% for GPT-6 Astra.
📊 It ranks first on the Vals Index, reflecting its high performance in legal and other knowledge work.
🔐 For now, only cybersecurity defenders in Google's Fairwind program can use it.
Full story in our weekly AI newsletter, The Batch 📬
#DeepLearningAI#AIEngineering#Cybersecurity
💻Learn by building with an AI coding agent, right in your browser. 👀Live preview, nothing to install.
📚Try new agentic coding lab in our course, Building AI Assistants with On-Device Memory.
👉Enroll for free:
#AI#AIAgents#AIEngineering
🧠 Yann LeCun helped invent the convolutional neural network and shared the 2018 Turing Award with Hinton and Bengio.
Now he joins us at AI Dev NYC to share his latest work at AMI Labs.
📍 NYC
🗓️ Nov 30 to Dec 1
👉
Just. Keep. Building.
📊 GPT-6.1 Sol: 52 on the AA Intelligence Index, at $0.72 per task.
GPT-6 Astra: 53, at $3.26 per task.
⏸️ OpenAI also canceled GPT-6.1 Astra’s release after the model showed deception in internal tests.
Full breakdown in The Batch 📬
#DeepLearningAI#AIEngineering#AIAgents
