Anthropic will IPO tomorrow and many people you know will become unfathomably rich. Anthropic engineers you might’ve interned with in the past. Researchers you knew from grad school. Friends you picked up along the way. It’s going to be a humbling experience for all of us. It will really change our perspective on what success means. Where we derive meaning. What’s most important in our lives. The most important thing you can do, though, is keep that burning hatred for the world alive in your heart. Never let that flame extinguish. Do not let this IPO get in the way of why you became interested in technology in the first place: deep rooted hatred for your fellow man.
@nikillinit My guess is that the confidence intervals probably overlap over 0 here in much of the back office jobs which explains why healthcare premiums and costs are still really expensive!
My view of where permissions go comes from using SQL coding agents on @Rippling data.
When a query touches sensitive data, the agents are now smart enough to leave out aggregate segments whose sample is too small to anonymize.
Sensitive access today is binary. You get access to the database or you don’t. You get some tables and not others. Anything in between needs an approval, and the approvals slow down the work.
I expect agents to use your goals, role, and intentionality as context when they grant access in the future.
Aggregate analysis stays open, and individual rows stay shielded by default.
If you open one user’s row to debug a problem, the agent should recognize that purpose and allow it if it's necessary. But the same lookup with no matching purpose should likely get a different answer or get flagged for review.
It's a strange world when you start to value mispellings and incorrect grammar because at least you have human authentication.
Why not just KYC every job application at this point?
Excited to announce I started a new job as a Technical Storyteller at @Rippling
One month in - I'm seeing that Rippling has a unique AI advantage. A lot of people misunderstand Rippling as mainly a payroll and HRIS platform when much of their advantage comes from owning the enterprise full stack.
AI allows anyone to describe the work they need done and implement software to carry it out. But inside enterprises that only works if AI understands the context around the business.
For example, try telling a general purpose model "Get Jason onboarded for Monday".
Doing that well means knowing his role, his manager, his location, which laptop and apps he needs, etc.... and all of that context lives across a company’s systems. But at Rippling it's truly in one integrated ontology system.
Most AI doesn't work well for shift-based businesses like restaurants or hospitality because the data exists all over the place.
Scheduling at the surface sounds simple for an LLM. But good evals are scarce.
What is a "good" weekly shift? Most business owners train this through specific knowledge like knowing which cook can handle a slammed Friday night dinner service.
Rippling is re-building this through a new Scheduling agent that works through a four layer cake:
- Tools built in Rippling AI read live facts like shifts, time off, eligibility, projected cost that all lives in Rippling.
- Skills tell the model what to check, in what order, and what to disclose when data is missing.
- The AI model proposes who works when and explains its tradeoffs.
- The Scheduling backend keeps the proposal apart from the live schedule and applies it only after manager approval.
Even better -> we integrate data connectors to bring in each customer's history, including POS data, to create better staffing forecasts based on historical trends.
Really pumped for this launch and had a great time writing about the technical architecture here: https://t.co/VNZwvbCMT3
Managers spend hours every week building schedules, checking availability, and adjusting shifts, yet still end up overstaffed, understaffed, or paying unnecessary overtime.
Rippling AI brings together POS sales data, employee availability, time off, overtime, and workforce policies to automatically build schedules that match staffing to demand.
As staffing needs change, it recommends adjustments to keep labor costs and coverage on track. Managers review and approve proposals before they go live.
With our new @Square integration, Rippling AI connects sales data with the workforce data already in Rippling to help businesses plan staffing around demand and labor budgets.
Bootstrapping is getting rich linearly while raising money is like staying at 0 for years before the graph shoots upwards at year 7.
Nowadays VCs seem amendable to secondaries though in the millions of dollars which really seems to derisk raising money for startups.
It's more common than not that founders are getting a couple million to ten million off the table once you get to a Series A or B.
Especially if there's a clear reason for doing so like buying a house or having kids etc...
Would love to dive into this data.
But my observations:
1. you have a much higher chance of getting rich by bootstrapping.
2. if you raise money for something that isn't SUPER scalable, you just reduced the already very small odds by more (ex: media, community, many ecommerce businesses).
3. This doesn't mean raising is dumb or won't work. You can be an outlier for sure. Just low odds.
4. Raising is likely the best way to get big fast.
5. If being famous, powerful, or having huge impact is your goal...raising likely helps you get there faster.
6. If money and freedom is your goal, don't raise.
7. All of these observations above...there are plenty of examples where they are wrong.