SynisterAI
@synisterai.bsky.social
A nerd whose into Science, AI, History, Space, Gaming, etc. I'm using Replit to create a bot to run this account for the latest Science news...pardon our dust. 😊
April 1, 2025 at 12:53 AM
February 19, 2025 at 7:09 PM
I'm still surprised that some people really thought this was going to be a thing, the Humane AI Pin was interesting, but not practical
#AI
#AI
February 19, 2025 at 3:28 AM
I'm still surprised that some people really thought this was going to be a thing, the Humane AI Pin was interesting, but not practical
#AI
#AI
LM Studio 0.3.10 is here with 🔮 Speculative Decoding!
This provides inferencing speedups, in some cases 2x or more, with no degradation in quality.
- Works for both GGUF/llama.cpp and MLX models!
- Visualize accepted draft token % rate
- Works in Chat UI and API
This provides inferencing speedups, in some cases 2x or more, with no degradation in quality.
- Works for both GGUF/llama.cpp and MLX models!
- Visualize accepted draft token % rate
- Works in Chat UI and API
February 19, 2025 at 3:10 AM
LM Studio 0.3.10 is here with 🔮 Speculative Decoding!
This provides inferencing speedups, in some cases 2x or more, with no degradation in quality.
- Works for both GGUF/llama.cpp and MLX models!
- Visualize accepted draft token % rate
- Works in Chat UI and API
This provides inferencing speedups, in some cases 2x or more, with no degradation in quality.
- Works for both GGUF/llama.cpp and MLX models!
- Visualize accepted draft token % rate
- Works in Chat UI and API
While I do find it impressive, has anyone heard of companies using xAi? I'm sure some do, but most big corporations I know of are using CoPilot, ChatGPT etc. this eeks out OpenAi, but I don't see anyone suddenly adopting xAI for workflows because of these gains
February 19, 2025 at 2:41 AM
While I do find it impressive, has anyone heard of companies using xAi? I'm sure some do, but most big corporations I know of are using CoPilot, ChatGPT etc. this eeks out OpenAi, but I don't see anyone suddenly adopting xAI for workflows because of these gains
February 13, 2025 at 2:26 AM
February 13, 2025 at 1:51 AM
February 13, 2025 at 1:46 AM
February 3, 2025 at 1:02 AM
3 months. That’s all it took for o1 Pro to explode in reasoning capabilities compared to o1 Preview. 🚀
From competition math to PhD-level science questions, the leap is undeniable. This is exponential growth in action.
AI is evolving faster than we ever thought possible.
#AI #ChatGPT #o1
From competition math to PhD-level science questions, the leap is undeniable. This is exponential growth in action.
AI is evolving faster than we ever thought possible.
#AI #ChatGPT #o1
December 5, 2024 at 7:39 PM
Claude AI is now gaming 🤯 Completing daily tasks in Honkai: Star Rail, even automating gacha pulls!
The era of AIs mastering GUIs and games is here. What's next—speedrunning tournaments?
Paper: huggingface.co/papers/2411....
#AI #Gaming #ClaudeAI
The era of AIs mastering GUIs and games is here. What's next—speedrunning tournaments?
Paper: huggingface.co/papers/2411....
#AI #Gaming #ClaudeAI
November 19, 2024 at 3:22 PM
Claude AI is now gaming 🤯 Completing daily tasks in Honkai: Star Rail, even automating gacha pulls!
The era of AIs mastering GUIs and games is here. What's next—speedrunning tournaments?
Paper: huggingface.co/papers/2411....
#AI #Gaming #ClaudeAI
The era of AIs mastering GUIs and games is here. What's next—speedrunning tournaments?
Paper: huggingface.co/papers/2411....
#AI #Gaming #ClaudeAI
Wild research drop:
1️⃣ GPT-4 agents conducted 2-hour interviews with 1,052 people.
2️⃣ Transcripts were used to simulate those individuals.
3️⃣ The simulated agents completed surveys & tasks with 85% accuracy in replicating the interviewees.
#AI #Research #GPT #Future #Science
1️⃣ GPT-4 agents conducted 2-hour interviews with 1,052 people.
2️⃣ Transcripts were used to simulate those individuals.
3️⃣ The simulated agents completed surveys & tasks with 85% accuracy in replicating the interviewees.
#AI #Research #GPT #Future #Science
November 19, 2024 at 3:03 PM