How to keep enjoying #programming in a world of #LLMs

How to keep enjoying #programming in a world of #LLMs
This seems too…naive and optimistic 🧐:
“If AI Coding Is Lowering Your Code Quality, You’re Not Managing Quality Right”, Iouri Khramtsov (www.i-kh.net/p/if-ai-codi...).
Via HN: news.ycombinator.com/item?id=4977...
#AI #LLMs #GenAI #AIAssistedCoding #SoftwareEngineering #Quality #Tests #Bugs
New research from Martian: picking the right model per prompt (not a fixed default) cut error rates 54% at the same cost across 44 LLMs -- 82% with multiple attempts. Worth testing before you assume your default model is still the right one.
arstechnica.com/ai/2026/06/a... - #AI #LLMs cost /a lot/ more now. It's more in line with their expenses, so more sustainable. But it's still mighty expensive.
Meet Graphify - an #opensource tool that turns codebases and unstructured docs into queryable #KnowledgeGraphs to optimize LLM workflows.
Recent updates have improved parser features & cross-file resolution.
🔗 Learn more via #InfoQ – buff.ly/nEOMJ1x
#AI searchers want answers instead of links
63% of US AI search users turn to AI assistants to answer a question directly
What is an LLM API key and how do you get one? healthaiinsights.substack.com/p/executive-... #llms #apikey
I'm glad the AI world is (re)discovering non-autoregressive models. LLMs are incredibly capable, but they aren't the right fit for every problem. If a faster, more energy-efficient model can solve a specific task, that's a clear win in my book.
Unlike traditional fixed-label classifiers, its label set is specified at inference time rather than training time. This structure gives it more flexibility.
Jev is a new model gaining traction. It's fast and cheap, designed to classify unstructured data instead of generating text. This shift back to non-generative methods is important for certain NLP tasks.
Large language models have the potential to enhance various aspects of autonomous driving, from perception to interactive decision-making. How do you design LLMs for autonomous driving? Here's an overview:
Jev's open-source alternative Laya-MLX runs on Apple Silicon with 1GB RAM and 7ms decisions. Could local decision models replace cloud LLMs for structured tasks?
#Jev #Laya #LayaMLX #AppleSilicon #LLMs
pythonlibraries.substack.com/p/jev-open-s...
Explore advances in:
✔ AI-powered molecular design
✔ Target identification & validation
✔ Drug repurposing
✔ Clinical translation
✔ Next-generation therapeutic development
🌐 ai-medicalcongress.com
#AIIM2027 #DrugDiscovery #GenerativeAI #LLMs
Another day, another talk (actually, the first of two today!) at Hong Kong University:
"LLM Math: Prospects and Challenges" by Boris Babic (HKU, Institute of Data Science and Department of Philosophy).
Agent 1st World! More than 50% of the web requests are now from AI subscriptions, agents or bots. To take advantage of this “Agent 1st World” your internet strategy needs to change – starting with your website. #llms.txt #Agents
www.youtube.com/shorts/ZTZ9...