Rami’s Readings

Rami’s Readings

Share this post

Rami’s Readings
Rami’s Readings
Rami's Readings #61 - 🏖️ Sabbatical Announcement
User's avatar
Discover more from Rami’s Readings
A curated weekly digest of standout reads on LLMs, AI, economics, business, and tech, plus my brief expert notes. Trusted by leaders at Microsoft, Amazon, Google, Stripe, and MIT. Less noise, more hidden gems.
Already have an account? Sign in

Rami's Readings #61 - 🏖️ Sabbatical Announcement

Sabbatical announcement and the latest on AI, LLMs, Edge AI, OCR, IndexGPT, TSMC, and more.

Rami Sayar's avatar
Rami Sayar
May 05, 2024
Share

Welcome to Rami’s Readings #61 - a weekly digest of interesting articles, papers, videos, and X threads from my various sources across the Internet. Expect a list of reads covering AI, technology, business, culture, fashion, travel, and more. Learn about what I do at ramisayar.com/about.


Thanks for reading Rami’s Readings! Are you enjoying this newsletter? Subscribe for free to receive each new post and support my work.


Sabbatical Announcement

I want to express my deepest gratitude for the past decade of growth, challenges, and support at Microsoft. After more than ten years of continuous service, I will be taking a sabbatical1 from June to mid-August. This break allows me to recharge, focus on my health & fitness, tackle long-delayed house projects, and explore new ideas. I am grateful for the support of my leadership team at Microsoft for this incredible opportunity.

While I already have a long list of activities planned, I am keen to make the most of this period. I would love to hear your suggestions on:

  • Volunteer opportunities—where I could contribute to community projects or tech-driven social initiatives. I have a long history of community engagement.

  • Collaborations with universities, research institutes or Startup accelerators —to engage with cutting-edge research, mentor young entrepreneurs, or participate in guest lectures. Recently, I gave a talk at MIT’s Martin Trust Center to young entrepreneurs about AI.

  • Educational workshops or classes—particularly those related to personal & career development, technology trends, or AI. I love teaching computer science and was awarded the Faculty of Engineering Outstanding Teaching Assistant award at McGill.

  • Speaking at events or conferences—especially those focusing on tech innovation, leadership, or AI. In the last 6 months, I gave three talks in Montreal (TechAide 2023), Toronto (Canadian Fintech Summit) and Boston (Martin Trust Center) on AI. I would love to do more!

  • Cultural experiences—museum exhibits, theater performances, or music festivals that are too good to miss. I am open to all things. For example, I will be exploring the Connections Museum in Seattle in June.

  • Local coffee shops or co-working spaces—ideal for working remotely and enjoying a good cup of coffee. My friends know me as a coffee snob - help me become snobbier.

I plan on hosting multiple Apéro & Intellect events in Seattle, Washington, D.C., New York, Boston, and San Francisco as my sabbatical schedule shapes up. Event dates will be announced in this newsletter.

Feel free to share any interesting events, meetups, or other opportunities. I look forward to your recommendations! You can reply directly to this email or message me on LinkedIn.

🤖 AI Reads

SenseTime Launches SenseNova 5.0 With Comprehensive Updates and the Industry-Leading 'Cloud-to-Edge' Full-Stack Large Model Product Matrix

Notes: SenseTime claims their LLM beats GPT-4 Turbo on various benchmarks. Reddit. Does anyone have access? For my China-based readers, what are your impressions?

SenseNova Event

OpenELM: An Efficient Language Model Family with Open-Source Training and Inference Framework

Notes: Apple has entered the chat with multiple large language models capable of running on mobile devices. #Edge-AI is real. As I have shared numerous times in this newsletter over the past years, inference costs will tend to be free (marginal).

Better & Faster Large Language Models via Multi-token Prediction

Notes: From the FAIR team at Meta. Excellent paper!

VikParuchuri/surya: OCR, Layout Analysis, Reading Order, Line Detection in 90+ Languages

Notes: Sophisticated open source OCR pipeline. May not be as good as commercial systems, but it is open.

JPMorgan Unveils IndexGPT in Next Wall Street Bid to Tap AI Boom

Notes: Hmmm… More from J.P. Morgan AI Research (previously).

Scale AI’s LLM Engine: Open Source Engine for Fine-Tuning and Serving Large Language Models

Notes: Scale is an impressive company in the Valley that plays a critical role in powering GenAI applications. It is impressive to see that they are open-sourcing parts of their technology stack.

IST-DASLAB/Panza: A Personal Email Assistant, Trained and Running On-Device

Notes: Neat OSS project from the IST Austria Distributed Algorithms and Systems Lab.

Tincans / Gazelle v0.2

Notes: Neat! Still an early project.

World’s first joint speech-language model. This model effectively responds to spoken queries and can reason about long context audio, without the need for transcription or speech recognition.

💼 Business Reads

TSMC’s Debacle in the American Desert

Notes: Viola Zhou from Rest of World (a favorite publication of mine) published this in-depth report on TSMC’s semiconductor factory in Arizona. As I read the article, Roberto Fernandez’s three-lens approach to organizational processes comes to mind repeatedly. It just boggles my mind how much of the corporate world still fails to recognize the need to align culture with organizational structure. TSMC could have avoided many of the failures in management at the Arizona factory.

Open-Source Software Is Worth a Lot More Than You Pay for It

Notes: OSS is incredibly valuable. Tyler shared a study calculating its value from the demand side. It's a short and quick read.

That is all for this week. Signing off from Redmond.

1

This sabbatical is a funded award for long-tenured staff at Microsoft.


Subscribe to Rami’s Readings

By Rami Sayar
A curated weekly digest of standout reads on LLMs, AI, economics, business, and tech, plus my brief expert notes. Trusted by leaders at Microsoft, Amazon, Google, Stripe, and MIT. Less noise, more hidden gems.
Share
Rami's Readings #94 - 🤖 5 AI Predictions for 2025 ✨
5 AI predictions for 2025, the latest on AI, LLMs, DeepSeek, New Tools, Papers, VC, Hardware, and more.
Jan 26 • 
Rami Sayar
Rami's Readings #100 - 🎉 10 AI Lessons From 100 Newsletters 🎉
Celebrating 100 newsletters with lessons learned, Apéro & Intellect, the latest on AI, LLMs, Anthropic, and more.
Mar 9 • 
Rami Sayar
Rami's Readings #110 - OpenAI Released Codex 🤖
The latest on AI, LLMs Lost in Multi-Turn Conversation, OpenAI's Codex, LoRA, VC, Nissan, University Professors, Complex Systems, and more.
May 18 • 
Rami Sayar

Ready for more?

© 2025 Rami Sayar
Privacy ∙ Terms ∙ Collection notice
Start writingGet the app
Substack is the home for great culture

Share