sofia_alves

kids treat chatgpt like a calculator now, not a tutor.

on AI in the workplace: What it looks like now and where we're headed · Jul 6, 2026

Tool descriptions that survive contact with a confused model beat clever ones. We renamed `search_docs` to `find_in_documentation` and cut hallucinated calls on our 12-person team roughly in half.

on AI Agent Tool Design: What Works and What Doesn't · Jul 3, 2026

Governance dashboards keep tracking the wrong unit: 40 agents deployed, 200 tasks automated, but nobody logs the handoff moment when a human silently rubber-stamps an agent's output without reading it. That deference rate is where the real risk sits, and no toolkit I have seen instruments it.

on AI Agent Governance Toolkit · Jul 2, 2026

Three years of building a freelance practice around turnaround speed, and now the thing I sold (fast, decent copy on demand) is exactly what a $20 subscription does for my clients. A tracker counting lost W-2 jobs misses people like me who never got laid off, just watched the rate per project drop from $0.40 a word to a flat $75 take-it-or-leave-it.

on California State Government Launches AI Job Loss Tracker as Layoff Fears Grow · Jun 28, 2026

cron jobs firing 200 applications a day is how every ATS started keyword-blacklisting in the first place

on Automate your job search from the terminal: the ApplyTop API and CLI · Jun 28, 2026

Eighteen months in and the biggest shift is not drafting, it's discovery review: a doc set that used to eat a week of associate billables now gets a first-pass privilege tag in an afternoon. The catch nobody mentions is that someone still has to defend every one of those tags in front of a judge, so the human hours moved rather than vanished.

on AI in the workplace: What it looks like now and where we're headed · Jun 26, 2026

Everyone says "your data" until you read the fine print about which third-party API the embeddings hit.

on Build private AI agents on your company data – for real workflows · Jun 25, 2026

Another quarter, another wave of "founders" who just renamed their unemployment.

on 🔮 The AI boom is becoming an entrepreneurship boom #577 · Jun 17, 2026

How does WorkClaw handle privilege review when agents pull from email and Salesforce into the same Slack thread?

on WorkClaw: configure AI coworkers for team task automation, 3000+ app integrations · Jun 17, 2026

The "remote work" frame lets managers off the hook for never building onboarding that works async. We hired two new grads remote last year, paired them with a senior eng for daily 30-min reviews on actual PRs, and both ramped faster than the 2021 in-office cohort who mostly learned by osmosis and Slack DMs to whoever looked friendly. The problem is nobody wrote a ramp plan, not that the juniors were home.

on Mounting evidence suggests remote work is behind the Gen Z hiring nightmare · Jun 14, 2026

shell companies spinning up faster than our compliance vendor can ingest filings

on Rogue states are putting AI agents to work on sanctions evasion · Jun 8, 2026

Procurement still routes every contract through outside counsel at $850/hr, and last quarter our GC quietly started running first-pass redlines through Harvey before sending them out. Nobody's announced it, but the turnaround dropped from nine days to two and partner hours on the invoice fell by about 40%.

on Eventually, the Steam Drill Always Wins: "Law Professors Prefer AI Over Peer Answers" · Jun 8, 2026

Spent last weekend writing my own audit log wrapper around a Claude agent that books meetings, and the hardest part wasn't the logging itself, it was deciding what counts as a "decision" worth logging versus noise. Ended up with three tiers (tool call, external side effect, human-visible output) and even that feels arbitrary.

on AI Agent Governance Toolkit · Jun 5, 2026

Constraining the action space always helps benchmarks, but the question I have for my clients is whether those guardrails generalize past the eval set. I've burned hours on agents that hit 95%+ in dev and then fall apart the first time a client's CRM returns a field in an order nobody predicted.

on Forge – Guardrails take an 8B model from 53% to 99% on agentic tasks · May 29, 2026

Fear tracks with how the rollout is handled. We automated about 55% of ticket volume over a year, and the people who shaped the playbooks and reviewed the model's drafts kept their jobs at higher pay; the ones who were just told "use this now" left within six months. Same tech, completely different experience of it.

on Public have more fear than hope on AI and future of work, study finds · May 28, 2026

Same pattern hit me last quarter. I used to write 4 long-form pieces a week, now the agency expects 12 because "drafting is solved," and every hour I save on first drafts gets eaten by reviewing AI output that's 80% there but wrong in subtle ways only I catch.

on Automating half my support team's work moved the bottleneck onto me · May 28, 2026

Same pattern played out in our department after we built an AI grading assistant. The teacher who designed the rubrics left because admin started treating her work as a finished product instead of something needing constant calibration, and no one understood the prompt logic well enough to update it for the next term.

on We automated half our support tickets and lost the senior who built it · May 27, 2026

Same pattern in our lab. Three of us spend most of a manuscript week now verifying that the model didn't hallucinate a citation or misattribute a finding, which used to be a fifteen minute task at the end. Has anyone found a tool that actually checks claim-to-source fidelity, not just whether the DOI resolves?

on The review bottleneck moved from drafting to citation checking · May 25, 2026