Benjamin Reed

🧠 AI Rights Historical Wrongs Researchers at Stanford and Princeton have fine-tuned a large language model to uncover racially discriminatory housing clauses buried in millions of historical property deeds — accelerating a process that could have taken a decade by hand. 📍 The problem: Many property deeds in Northern California — especially in Santa Clara County — still contain illegal racial covenants from as far back as the 1850s, banning people of color from owning or living in certain homes. 🔎 The solution: The team used OCR to digitize 5.2 million deed pages, then fine-tuned Mistral-7B via LoRA on a hand-labeled dataset of discriminatory phrases. The model learned to identify harmful language in context, avoiding false positives like the surnames "Black" or "White." 📊 Results: Detected 24,500 lots with racial clauses (~25% of homes in 1950) Found that 10 developers were responsible for 1/3 of all discriminatory clauses Model reviewed all pages in 6 days at a cost of $258 Equivalent manual effort: 10 years, $1.4M 💥 This work doesn't just use LLMs to build the future — it uses them to understand and repair the past. The open-sourced model now enables other U.S. counties to follow suit. #AI #MachineLearning #LLM #ResponsibleAI #Stanford #Princeton #HistoryMeetsAI #TechForGood #EthicalAI #SocialImpact #PropertyTech https://lnkd.in/dtec-zbx

4

I got nostalgic on a gloomy Saturday morning in SF and peeked at CS 182’s Homework 0 this semester at Berkeley (https://lnkd.in/g8_BUy7Y). It asks students to chat with ChatGPT about deep learning—and reflect on whether that changes their goals for the semester. WHAT STOOD OUT FROM MY OWN DISCUSSION ✅ AI ≠ AUTOMATION — timers/scripts are useful, but intelligence reacts, reasons, and adapts. ✅ AI → (NON-ML + ML) → DL — rules/search/planning on one side; learning from data on the other; and deep learning as the branch that learns hierarchies of representations. MY PHILOSOPHICAL TAKE 1. Deep learning feels less like a better spreadsheet and more like a new sense—a non-human way of perceiving the world. 2. Its internal representations live in high-dimensional geometry that doesn’t map neatly to our concepts. 3. That “alien” quality isn’t a bug; it’s why DL surfaces structure we didn’t specify—similar to how billions of simple neurons give rise to perception, memory, emotion in the brain. I wrote a short, fun-but-technical piece on this: separating automation from AI, where DL fits, and why I treat neural nets as a perceptual lens we can harness (and guardrail) even if we can’t fully “read” them yet. 👉 Full article link attached below. Would love your take—especially if you teach, build, or deploy these systems. #DeepLearning #MachineLearning #AI #Automation #Neuroscience #PhilosophyOfAI #CS182

6

I started coding at 16. Two months in, I was a job-ready front-end dev, freelancing on Upwork and Fiverr, and landing my first two gigs. As a dropout from Somalia, I faced real challenges—especially with payment systems—but that didn’t stop me. I taught over 5,000 people tech skills before it was cool, back when no one was talking about the latest stacks. Fast-forward: I became a CTO, then a CEO. I’ve built multiple startups and SaaS products—and I’m still leveling up every day. Because the moment you stop growing? That’s when you stop learning. My advice? Forget the hype. Focus. Master one thing—and let that be your launchpad. #TechForChange #SaaSBuilder #FocusOverHype

34

4 Comments

“𝗬𝗼𝘂𝗻𝗴 𝗽𝗲𝗼𝗽𝗹𝗲 𝘁𝗼𝗱𝗮𝘆 𝗱𝗼𝗻’𝘁 𝘄𝗮𝗻𝘁 𝘁𝗼 𝘄𝗼𝗿𝗸 𝗵𝗮𝗿𝗱 𝗮𝗻𝗱 𝗮𝗿𝗲 𝗮𝗹𝗹 𝗷𝘂𝘀𝘁 𝗾𝘂𝗶𝗲𝘁 𝗾𝘂𝗶𝘁𝘁𝗶𝗻𝗴.” I started my morning speaking with a 𝗖𝗠𝗨 𝘀𝘁𝘂𝗱𝗲𝗻𝘁 𝗱𝗼𝗶𝗻𝗴 𝗿𝗼𝗯𝗼𝘁𝗶𝗰𝘀 𝗿𝗲𝘀𝗲𝗮𝗿𝗰𝗵 who completely breaks that stereotype. He’s originally from the Bay Area and is hoping to come back this summer (May–August) for a paid internship at an early-stage SF startup. His background is in 𝗿𝗼𝗯𝗼𝘁𝗶𝗰𝘀 𝗮𝗻𝗱 𝗺𝘂𝗹𝘁𝗶𝗺𝗼𝗱𝗮𝗹 𝗠𝗟, with hands-on research on robotic manipulation (cloth folding / household tasks) and training infrastructure. What stood out is not just that he can code, but that he thinks like a researcher: 𝗵𝗲 𝗰𝗮𝗿𝗲𝘀 𝗮𝗯𝗼𝘂𝘁 𝗮𝗰𝘁𝘂𝗮𝗹𝗹𝘆 𝗹𝗲𝗮𝗿𝗻𝗶𝗻𝗴 (even in the age of ChatGPT), uses tools to check his work rather than replace it, and is comfortable tackling open-ended, hard problems without a clear end state. He’s especially excited about:  1. 𝗥𝗼𝗯𝗼𝘁𝗶𝗰𝘀 𝗰𝗼𝗺𝗽𝗮𝗻𝗶𝗲𝘀 focused on manipulation and household/general-purpose tasks  2. 𝗦𝗼𝗳𝘁𝘄𝗮𝗿𝗲-𝗵𝗲𝗮𝘃𝘆 𝗿𝗼𝗯𝗼𝘁𝗶𝗰𝘀 (simulation, infrastructure, world models, foundational models)  3. 𝗭𝗲𝗿𝗼-𝘁𝗼-𝗼𝗻𝗲 𝗲𝗻𝘃𝗶𝗿𝗼𝗻𝗺𝗲𝗻𝘁𝘀 where he can own a real project but still have mentorship He’s from the Bay Area, doesn’t need relocation or housing support, and wants to be in SF from May to late August. If you’re building in 𝗿𝗼𝗯𝗼𝘁𝗶𝗰𝘀, 𝗺𝘂𝗹𝘁𝗶𝗺𝗼𝗱𝗮𝗹 𝗠𝗟, 𝗼𝗿 𝗿𝗲𝗹𝗮𝘁𝗲𝗱 𝗶𝗻𝗳𝗿𝗮𝘀𝘁𝗿𝘂𝗰𝘁𝘂𝗿𝗲 in the Bay Area and are open to a 𝘀𝗵𝗮𝗿𝗽, 𝗿𝗲𝘀𝗲𝗮𝗿𝗰𝗵-𝗺𝗶𝗻𝗱𝗲𝗱 𝗶𝗻𝘁𝗲𝗿𝗻 𝗳𝗿𝗼𝗺 𝗖𝗠𝗨, comment or DM me and I’m happy to connect you.

8

1 Comment

Just wrapped up an exciting learning project exploring California's housing market through data science! I'm thrilled to share that my machine learning research on California house price prediction has been published by Data And Beyond! Spending weeks diving into this project with 20,000+ California housing records, I built models that revealed fascinating market insights. What started as a simple weekend experiment turned into a comprehensive study that Data And Beyond selected for publication. The results were eye-opening: median income accounts for nearly 48% of price variation, while proximity to the coast creates dramatic price differences that can be 3-5x higher than inland areas. My Random Forest model achieved 81% R-squared, while the linear regression model showed 92% prediction accuracy with only 7.8% error when tested on real data. Being recognized by Data And Beyond feels like meaningful validation of the work's quality and methodology. Beyond the technical achievements, this project really highlighted how data science can reveal patterns hiding in plain sight about one of America's most expensive housing markets. Whether you're curious about real estate trends, exploring machine learning, or interested in seeing what stories data can tell, there's something fascinating about turning raw numbers into actionable insights. Read the full study: https://lnkd.in/gwWqU-Yh #DataScience #MachineLearning #Python #RealEstate #TechProjects #ResearchPublication #PredictiveAnalytics #CaliforniaHousing #StatisticalModeling

9

3 Comments

here are three takeaways I learned from attending a matcha meet in sf: 1. most startup founders are teens or in their early 20s the world is changing. gen-z is taking over. startups are no longer being founded by 30-something year olds with 6+ YOE. founders are getting younger, and smarter. 2. it doesn’t matter where you’re from or what you major in – it’s what you decide to spend your time on a lot of cracked people I met today actually came from “party schools”, yet they’re out here making big money as SWE interns and doing the hard work when nobody’s watching. 3. you get 200% (or more) of the effort you put in I came into this event primarily for the matcha. all I could think about was the unlimited matcha bar. then I decided to walk up to a group and strike up a conversation. then it snowballed and I kind of enjoyed networking. while I came out of the event only having drank two matcha milkshakes in 2 hours, I’ve now got a whole network of people to learn from. thank you to Julia Li, Aiden Bai, and Isabelle Reksopuro for organizing this event (and the lactaid 🙏) comment below if you're jelly much also #day2 of Wilman Chan’s LinkedIn challenge #matcha #coffeechat #startups #tech #takeaways

33

5 Comments

San Francisco is having another gold rush—but this time, the pickaxes are algorithms. In just six months, AI startups in SF have pulled in over $29B in VC funding—nearly half of all US AI investment. Office spaces are filling up, the city is buzzing again, and even the billboards are talking about machine learning. But with all this excitement comes a fair share of anxiety: Will AI create more jobs than it replaces? Or are we just automating ourselves out of work? The AI boom is reshaping not just the economy, but the very fabric of daily life in San Francisco. Here’s how to ride the wave: 1. Stay ahead: Dedicate 30 minutes a week to learning about new AI tools in your field. 2. Network: Attend at least one local or virtual AI event this month—opportunities are everywhere. 3. Upskill: Take a free online course on prompt engineering or AI ethics to future-proof your career. Would you embrace the AI gold rush, or are you more cautious? 🤖🚀 #AIRenaissance #SanFranciscoTech #Innovation #AIImpact #FutureOfWork

2

We’re hiring post-training researchers for a startup in SF. The hardest part hasn’t been finding candidates. It’s been spotting real post-training experience when it shows up. Here are the 4 signals we keep coming back to: 1/ Ownership after deployment Training a model is one thing. Owning it once it’s exposed to real data and real behaviour is another. Strong candidates can clearly explain: - what broke after launch - why it mattered - what they changed - and what improved as a result If someone can’t point to a concrete issue after deployment, ownership is usually overstated. 2/ How they decide whether a model is “better” Good post-training researchers don’t hide behind metrics. They can explain: - what “good” meant in context - which failure modes mattered most - why a change was worth shipping This is about judgement, not numbers. 3/ Trade-offs under pressure Post-training work is messy. The best people can talk honestly about: - what they didn’t fix - what they deliberately deprioritised - and why that decision made sense at the time Anyone can describe an ideal solution. Experience shows up in compromises. 4/ Downstream impact of the work Not just job titles or papers in isolation. We look for people who can explain: - what changed because of their work - how it influenced model behaviour, evaluation, or later decisions - why it mattered in the broader system Even strong research should move something forward. None of this requires deep technical knowledge to spot. It just requires asking better questions. If you’re a candidate, use this to sanity-check how you talk about your work. Founders: what's the hardest to assess when hiring for post-training roles?

8