Sign In

Albert Chen

Monitoring & Evaluation Manager, Urban Mosaic

Albert Chen is an AI strategist and prompt engineer with extensive experience in startups, corporate environments, and nonprofits. He served as the Monitoring and Evaluation Manager at Urban Mosaic / Mosaico Urbano from July 2017 to September 2019, based in the Mexico City Metropolitan Area.1

After his role at Urban Mosaic, Chen became the Director of Administration & Finance for the same organization from September 2019 to June 2022.1 His experience also includes:

  • Co-founder at Anago Venture Builder (July 2023 - Present)
  • R&D at a Stealth Startup (January 2023 - Present)
  • Various roles at State Street, including Assistant Vice President of Technology
  • Marketing positions at AgaMatrix, Inc.

Chen has 18+ years of experience in the US and Latin America, specializing in merging his expertise with generative AI tools, large language models (LLMs), administration, and operations.2 He helps executives design and implement AI roadmaps and provides AI training for both technical and non-technical teams.2

Currently, Chen is focused on AI strategy and prompt engineering, leveraging his diverse background to assist organizations in their AI initiatives and digital transformations.12

Highlights

Mar 8 · linkedin

⛔ RAG isn't ALWAYS better! ⛔ Perplexity vs ChatGPT vs Claude vs Pi vs You.com vs Gemini vs Exa

I stumbled upon an excellent use case while searching for a place to celebrate my 10th anniversary last summer. Boston is typically hot, humid and even rainy, so I made sure to include "climate-controlled" in my prompt.

The response from Perplexity was shocking, hilarious and fascinating as I tried to understand why it responded as it did. Poor LLMs trying so hard to give us what we ask for, but not what we actually want... 😅

Apparently, I should have included the word "Romantic" ❤️

I ended up prompt engineering to get better results, giving me an amusing case study to share.

However, I realized this would also be an excellent way to test different LLM's abilities to intuit what I actually want, rather than just semantic search. So for this evaluation I kept the SAME PROMPT to see how different models would respond.

I've posted screenshots of each one with my rating and a few annotations:

Perplexity w/ RAG (1st try): F (all horrible answers) Perplexity w/ RAG (2nd try): D+ (1st answer still horrible) Perplexity w/ RAG (3rd try): B- (better answers, but 1st link still bad) ChatGPT 3.5: B+ (Good answers) - Free version ChatGPT 4: A+ (GREAT answers) - Paid subscription Claude 2.1: A (Great answers) - Free Claude 3 Opus: A (GREAT answers, but 1 error) - Paid Inflection AI Pi: B+ (Good answers) You.com Smart Mode: B+ You.com Research Mode w/ RAG: A (Great answers, but still a bit venue focused)

At the end of the day... Context ≠ Intent

These bots can anticipate the context and pull in a bunch of extra info via RAG, but that's not what I actually wanted.

Prompt Engineering will still exist for quite some time because HUMANS don't always know how to articulate what they actually want... and MACHINES can't read our minds... yet. 😅

#perplexityai #chatgpt #claude #pi #you

Related Questions

What specific projects has Albert Chen managed at Urban Mosaic / Mosaico Urbano?
How did Albert Chen transition from a career in finance to a role in project management?
What are some notable achievements of Albert Chen in his career?
How has Albert Chen's experience at State Street influenced his work at Urban Mosaic / Mosaico Urbano?
What skills does Albert Chen bring to his role as Monitoring & Evaluation Manager?
Albert Chen
Albert Chen, photo 1
Albert Chen, photo 2
Add to my network

Location

Mexico