Jay Guthrie
@StraughterG
Followers
1K
Following
24K
Media
395
Statuses
21K
Machine Learning Expert | Specializing in GPT Development & AI-Powered Systems | Innovator in Agent-Based Modeling | Advancing AI with Robust RAG Pipelines.
The United States
Joined January 2012
My career advice in one image
1
0
4
RT @fofrAI: Here's a new model I've pushed to Replicate. It uses gpt-5 with tool-use and ffmpeg to run prompt based video edits:. https://t….
0
19
0
🧠 Apollo MCP plus Dagger equals natural language control of container operations, builds, and CI/CD through Claude Code.
0
0
1
i have a m2 32gb.
Five neural nets, achieving completely local voice AI, no internet, on an M1 with only 16GB ram. Neural-based voice activity detection and turn detection means it's interruptible, but never interrupts me, and is able to sit idle and waiting. It's been flawless so far. 12B
1
0
1
RT @joshwhiton: Five neural nets, achieving completely local voice AI, no internet, on an M1 with only 16GB ram. Neural-based voice activi….
0
273
0
RT @raw_works: i think @DSPyOSS needs a "pruning mode". @LakshyAAAgrawal this might be easier than your envisioned prompt optimization +….
0
4
0
RT @MaximeRivest: I still need to triple and quadruple check, but it seems like we can indeed, sometimes, load only 10% of an MoE model for….
0
13
0
wow.
450M vision model runs at 353 tokens/sec on MLX using just 1.456GB memory 🔥.LiquidAI drops LFM2-VL-450M → MLX-VLM community with @Prince_Canuma. ports it → Your Mac becomes a vision AI beast.Liquid neurons adapting computation on the fly. 450M params punching way above its
0
0
0
RT @DittmannAxel: 450M vision model runs at 353 tokens/sec on MLX using just 1.456GB memory 🔥.LiquidAI drops LFM2-VL-450M → MLX-VLM communi….
0
18
0
RT @multimodalart: IT'S OUT! 🚀 MoDA: Multi-modal Diffusion Architecture for Talking Head Generation. finally a talking head:.open source 🏋️….
0
61
0
RT @SuperagenticAI: 🚀 Meet GEPA — the DSPy-powered optimizer in SuperOptiX that lets AI agents self-reflect and evolve smarter, faster. Les….
0
2
0
hmmm this can be useful.
perfect AI copywriting comes down to these five elements. ICP profile + system instructions + brand voice + business profile + markdown examples. here's what each piece does:. → ICP psychological mapping (decision triggers, not demographics).→ business intelligence files.
0
0
0
RT @melbaldove: Hi @lateinteraction. Whats a reasonable LLM as judge accuracy score to strive for? Im optimizing with GEPA using GPT-5 have….
0
2
0
The optimal communication protocol between world-understanding and action-taking is itself a complete specification of intelligence for that domain.
0
0
0
dogfood it again this time think outside the box for different aspects.
0
0
0
Every new alignment paper is basically: “make it less chaotic” 🤣.GEPA, RewardAgent, ActPRM… all entropy compressors 🗜️.Less noise, more coherence 🔊➡️🎵.
0
0
0
github.com
Powertools for livebook.dev — AI Code Editing, MCP Servers, and Running Livebooks from the CLI - thmsmlr/livebook_tools
Is there an LLM integration for Livebook that will write/edit cells?. #ElixirLang.
0
0
0
elixirlang.
Will the LLMs choose @elixirlang and @elixirphoenix given the ideal use-case for that stack?. Gemini & GPT-5: Yes.Claude 4 Sonnet & 4.1 Opus: No. #ElixirLang
0
0
1