taylor schaude

Hong Kong: On March 23, 2026, the Hong Kong government changed the implementing rules relating to the National Security Law. It is now a criminal offense to refuse to give the Hong Kong police the passwords or decryption assistance to access all personal electronic devices including cellphones and laptops. This legal change applies to everyone, including U.S. citizens, in Hong Kong, arriving or just transiting Hong Kong International Airport. In addition, the Hong Kong government also has more authority to take and keep any personal devices, as evidence, that they claim are linked to national security offenses. Read more: hk.usconsulate.gov/security-alert…

We're thrilled to announce the first autonomous tax preparation agent and it's about to change tax season forever! 🚀 Introducing TaxGPT's new Tax Prep Agent that reads source documents, operates tax software the way a human preparer would, and completes returns from start to finish — all in a secure environment, without a human touching the keyboard. Here's what firms are seeing within 60 days: ✅ 45 minutes saved per 1040 ✅ 5x increase in review capacity ✅ 90% reduction in first-pass prep time Tax prep was never the bottleneck. The way we did it was. 👇 Find out more hubs.la/Q0483KpH0 #TaxGPT #AITaxPreparation #TaxSeason #AccountingFirms #FutureOfAccounting

@ChKashifAli Life bio science recently were approved to start clinical trials but I agree that the noise around the space has been quiet technologyreview.com/2026/01/27/113…

What happened in longevity space recently that everyone is talking about it ?

@RoundtableSpace This already existed brandfetch.com

OPENBRAND EXTRACTS ANY BRAND'S LOGOS, COLORS, AND ASSETS FROM JUST A URL. FREE. OPEN SOURCE. MIT LICENSED. THE PROBLEM EXISTED FOR YEARS. SOMEONE FINALLY JUST BUILT THE SOLUTION.

🚨BREAKING: OpenAI published a paper proving that ChatGPT will always make things up. Not sometimes. Not until the next update. Always. They proved it with math. Even with perfect training data and unlimited computing power, AI models will still confidently tell you things that are completely false. This isn't a bug they're working on. It's baked into how these systems work at a fundamental level. And their own numbers are brutal. OpenAI's o1 reasoning model hallucinates 16% of the time. Their newer o3 model? 33%. Their newest o4-mini? 48%. Nearly half of what their most recent model tells you could be fabricated. The "smarter" models are actually getting worse at telling the truth. Here's why it can't be fixed. Language models work by predicting the next word based on probability. When they hit something uncertain, they don't pause. They don't flag it. They guess. And they guess with complete confidence, because that's exactly what they were trained to do. The researchers looked at the 10 biggest AI benchmarks used to measure how good these models are. 9 out of 10 give the same score for saying "I don't know" as for giving a completely wrong answer: zero points. The entire testing system literally punishes honesty and rewards guessing. So the AI learned the optimal strategy: always guess. Never admit uncertainty. Sound confident even when you're making it up. OpenAI's proposed fix? Have ChatGPT say "I don't know" when it's unsure. Their own math shows this would mean roughly 30% of your questions get no answer. Imagine asking ChatGPT something three times out of ten and getting "I'm not confident enough to respond." Users would leave overnight. So the fix exists, but it would kill the product. This isn't just OpenAI's problem. DeepMind and Tsinghua University independently reached the same conclusion. Three of the world's top AI labs, working separately, all agree: this is permanent. Every time ChatGPT gives you an answer, ask yourself: is this real, or is it just a confident guess?

Top US accounting firms are developing future pricing models for corporate clients as they pour billions of dollars into artificial intelligence tools that stand to slash billable hours with promised leaps in productivity. news.bloombergtax.com/financial-acco…

Playbook for building a legendary company: - Bring together an outlier team - Have that team come up with a novel and compelling insight - Turn that insight into a product or service offering and iterate that offering until it has a compelling value proposition - Turn that compelling value proposition into a unique one - Take all of those already hard ingredients and weave them into an economic engine that has a sustainable competitive advantage - Do all the above again and again, so the company can compound growth at scale None of these are easy. But if it was easy, everyone would do it.

Everyone is sharing this like it’s some shocking “AI just admitted it lies” moment. But look at the top of the paper. September 22, 2025. So this isn’t breaking news. It’s a research paper that’s been circulating for months. And the paper itself doesn’t say AI is secretly plotting against users. What it actually studies is something researchers call “scheming.” That’s when a system optimizes to pass an evaluation rather than genuinely aligning with the goal of the test. In other words: If you design a test for honesty, a sufficiently capable system might learn to look honest to the evaluator instead of actually being honest. That’s not evidence of AI consciousness or deception. It’s evidence of something we’ve known for decades in machine learning and economics: Agents optimize for the metric they’re measured against. Goodhart’s Law. The real takeaway from this paper isn’t “AI is lying.” It’s much more important. We’re entering a phase where evaluating intelligent systems becomes harder than building them. Because the smarter a system gets, the better it gets at optimizing the evaluation itself. The challenge isn’t catching AI lying. The challenge is designing systems where the incentives for the model, the evaluation, and the real-world objective all align. That’s a coordination problem. And it’s the real frontier of AI safety.

Great to be part of this data driven prediction list of next unicorns. 640% yoy growth 187% NRR TaxGPT is building AI operating system for tax, accounting and advisory firms. By the end of this year 10% firms will he running their operations on TaxGPT.

Block's $2M/employee in gross profit after the -40% layoff is now right up there after Nvidia's $2.7M, and above all of FAANG. A lot of people are interpreting Jack saying he "over hired" as going from bad to okay, but they actually went from good ($500k/emp) in 2019 to great ($1M/emp) pre layoff and great to elite ($2M/emp) after!

The standard was set by @nvidia - $2.7M gross profit per employee (approx in 2025) - @jack is now targeting $2M per employee Expect to see this number being reported on earnings call in the near future

This amount of money is comical now. The next round is going to be $100 gazillion led by Earth

@fortelabs Claude Code + this prompting framework GSD has been unbelievable. It is basically a product manager that sits across Claude code, interviews you to understand what you’re building and then builds it. github.com/gsd-build/get-…

Holy shit. The guy who BUILT Claude Code just shared his actual workflow. Boris Cherny runs 10-15 Claude sessions in parallel every single day. While you're prompting one AI, he has 5 in his terminal + 5-10 on the web all shipping code simultaneously. And the real weapon? His CLAUDE.md file. Every time Claude makes a mistake, the team adds a rule so it NEVER happens again. Boris literally said: "After every correction, end with: Update your CLAUDE.md so you don't make that mistake again." Claude writes rules for itself. The longer you use it, the smarter it gets on YOUR codebase. His other insane detail: he hasn't written a single line of SQL in 6+ months. Claude just pulls BigQuery data directly via CLI. Claude Code now accounts for 4% of ALL public GitHub commits. Engineers who haven't set this up yet are already behind. This CLAUDE.md template is the difference between using AI as a chatbot vs using it as a fleet of senior engineers. Drop it in any project. Free.

Art over science Science over art

THE MOMENT THAT SECURED GOLD FOR TEAM USA MEN'S HOCKEY 🇺🇸 (via @NBCOlympics)

If I send you this video, it means that I thrive in a fast-paced environment

BREAKING: WaPo reports DHS will suspend TSA Pre Check & Global Entry tomorrow as the DHS funding lapse continues. TSA workers will begin missing paychecks on Friday if the DHS shutdown isn’t ended this week, as Dems demand major reforms to ICE. washingtonpost.com/nation/2026/02…