Alex π PromptLeo
@alexzender
Followers
526
Following
771
Media
23
Statuses
286
Founder @PromptLeo, @AskDynaPictures. Prev. tech lead @oracle, tweets about AI, SaaS, marketing
Joined April 2009
Everyone is talking about MCP. Here's why you might NOT want to use it π«. First, what is MCP? Model Context Protocol (from Anthropic) is a standardized way for LLMs to detect actions in prompts that can be delegated to external tools, call those tools, get responses, and show
1
0
0
RT @markvitals: When will @OpenAI #GPT4o advanced image capabilities be released? Interested in #visualNarrative and consistent character/sβ¦.
0
6
0
A bit of AI humor on Friday π . Why did the Generative AI go to therapy?. Because it had too many unresolved issues and couldn't stop generating complex characters!. #GenerativeAI #humour #FridayFeeling
3
0
9
Scientists recently unveiled a generative AI that has mastered time travel through simulations. βοΈ. It creates detailed historical events that havenοΏ½οΏ½οΏ½t occurred, offering historians insights into alternate timelines where, for instance, the internet was invented in the 18th
1
0
34
RT @mprokopov: Andrew Ng, the person in AI not needed to introduce, shared in the post the patterns on how to improve LLM output. Namely: Rβ¦.
deeplearning.ai
The Batch AI News and Insights: Last week, I described four design patterns for AI agentic workflows that I believe will drive significant progress...
0
1
0
π°π°π° Happy Easter Prompts π°π°π°.#PromptShare #DALLE3 #SDXL . Generated Easter rabbits with DALLE (first two images below) and Stable Diffusion (third and forth). Prompts for DALLE πππ:. 1. Adapt the photorealistic scene by portraying the rabbit standing upright on its
1
0
5
Check out this new prompt engineering trick: βhere is the most relevant sentence in the context:β. It was probably used when training and tuning the model.
How do you increase a 27% accuracy score to 98%?. By adding βHere is the most relevant sentence in the context:β to your prompt. Yesterday, Anthropic published a blog post outlining how they achieved significantly better results on the same evaluation by adding this sentenceΒ to
0
0
2
RT @mustafasuleyman: 3 different people have sent me this video in under 24hrs, so I watched it and can confirm this is a VERY clear, accesβ¦.
0
221
0
Prompt engineering gets even more important as the context length increases. Cloude 2.1s performance pattern is similar to GPT-4. See the summary of the Cloude 2.1 test with 200K context below π.
Claude 2.1 (200K Tokens) - Pressure Testing Long Context Recall. We all love increasing context lengths - but what's performance like?. Anthropic reached out with early access to Claude 2.1 so I repeated the βneedle in a haystackβ analysis I did on GPT-4. Here's what I found:
0
0
3
Learn Generative AI in 12 lessons for free. Free course from Microsoft available on GitHub π.
github.com
21 Lessons, Get Started Building with Generative AI π https://microsoft.github.io/generative-ai-for-beginners/ - microsoft/generative-ai-for-beginners
Microsoft launched the best course on Generative AI. The free 12 lesson course is available on Github and will teach you everything you need to know to start building Generative AI applications. Each lesson includes:.- a short video introduction to the topic.- a written lesson
0
0
0
RT @alexzender: βοΈ βοΈβοΈ How to (re)write marketing texts in fairy tale style?. π§ββοΈπ§ββοΈπ§ββοΈ.Fairy tales traditionally use simplistic proseβ¦.
0
1
0