{{first_name | Hey}}, welcome back.
Today’s issue covers AI industry’s push to train a new generation of engineers, Wall Street’s surging demand for one AI skill, and Apple tightening the rules around AI agents accessing your data.
These are today's updates.
Anthropic puts $100M behind a new generation of AI engineers
Wall Street demand for agent orchestration jumps 1,721%
Apple moves to tighten Mac controls as AI agents gain access to more data
Tools, jobs, resources, and last issue’s winning prompt ⬇️
Tutorial
This 53-minute Grok Bot workshop from SpaceXAI’s GTM team (featuring experts like Krista Letz and Vincent Zhu) shows how to build and run a specialized team of bots. It covers clear role assignment, inbox/meeting prep automation, prospect research, outreach, and parallel task handling. It also shows how to onboard bots like employees using a simple role → tools → skill → feedback → routine framework.
Neatprompts is where enterprise teams discover what's worth paying attention to in AI. Every week, 100,000+ readers rely on us for actionable insights and practical workflows.
If your product helps businesses adopt, deploy, or scale AI, we'd love to introduce it to our audience.
Top News
Anthropic is putting $100 million into the Claude Frontier Academy, which aims to train 10,000 frontier deployed engineers by the end of 2027. The first cohorts come from Accenture, Bain, Capgemini, Deloitte, McKinsey, Morgan Stanley and Novo Nordisk. Each engineer does a multi-day in-person program, then spends 12 weeks leading a real Claude project at their own employer, with Anthropic engineers alongside. Credentials start in early 2027.
The news comes at a time when the biggest vendors race to put engineers inside customers. AWS committed $1 billion to forward-deployed engineering on June 30. Microsoft followed with a $2.5 billion Frontier Company built around 6,000 of its own staff.
Anthropic is training the people who already work at the consultancies and customers. Entry is by nomination through your Anthropic account team, and every nominee needs a named project. The New Stack analysis highlights some points worth thinking about first for those wanting to sign up.
AI job postings at JPMorgan Chase, Citigroup, Capital One and other big banks are up 49% this year, to 139,819, according to hiring firm Draup's analysis. The fastest riser is agent orchestration, up 1,721% to 1,967 mentions, still about 1.4% of listings. Governance leads by a wide margin at around 16,000 mentions, almost double the roughly 8,400 for building and running models.
Built by Neatprompts from Draup's job-posting data, as reported by CNBC. Sources: CNBC article 2 Oct 2026, Yahoo Finance report, Tech Startups coverage.
Federal model risk guidance issued in April, SR 26-2, may explain the emphasis. Its footnote says generative and agentic AI are outside its scope and tells banks to rely on their own governance practices. That leaves each bank writing its own rules for who approves what an agent can touch and when a human steps in, and the job listings read like a team being staffed for exactly that.
Apple is planning new limits on Full Disk Access, the Mac permission that lets an app read files, mail, Messages and browsing history. In a developer notice Friday, Apple said some developers use it in ways that expose users' data without full understanding, and that granting it will take very explicit user action. It gave no date or macOS version.
The notice follows Inc. columnist Jason Aten's claim that Meta's Muse agent synced 187,462 rows of his Messages with the permission switched off. Meta disputes it and says Muse needs both Full Disk Access and its Messages connector. MacStories says Apple's backup-app rationale ignores how many utilities hold it. Apple also warned that granting access to a messaging app can expose the people on the other side of those chats. Supervised Macs can already manage the permission through a device-management profile, and Apple hasn't said how the new rules will treat those grants.
Signals
Tools
ChatGPT Space: Create, collaborate, and build together with AI in Space
Calnio: Two-way sync between Notion and Apple Calendar
Miso: Book flights and hotels through iMessage
Resources
Poll
Choose a prompt you want us to publish next
Market
Funding
General Intuition raised $220 million to build foundation models trained on gameplay video
Supabase raised $150M to build infrastructure for AI-generated databases
Roles in AI
Technical Product Marketing Manager at Bloomreach
Chief of Staff, AI Engineering at Lenovo
Podcast
How do you transform a company with AI when you’re a non-technical CEO? Most CEOs delegate it to IT. Manuel Diez, CEO of Grupo Diesco, took the opposite approach, becoming his company’s AI leader and transforming the business in six months. In this episode with Geoff Woods, he shares what changed, what didn’t, and the simple discipline he used to turn AI into an execution system.
Prompt of the Day
AI Investment Portfolio Reviewer
When to use this?
Use this when you're reviewing multiple AI initiatives for quarterly portfolio reviews, budget planning, or preparing an AI investment update for leadership.
Act as an AI portfolio reviewer for an enterprise. Help me assess whether our current AI investments are delivering enough value and where we should increase, reduce, or redirect investment.
I’ll provide information about our AI initiatives, including use case, business owner, users, annual cost, adoption, expected benefits, current results, implementation status, and strategic importance.
For each initiative, evaluate:
Business value: What measurable value is it creating?
Adoption: Are people actually using it, and how deeply?
ROI: How does the realized value compare with the cost?
Strategic fit: Does it support a meaningful business priority?
Scalability: Can it expand across teams, regions, or workflows?
Risk: What are the biggest security, compliance, reliability, or vendor risks?
Then classify each investment as Scale, Optimize, Watch, or Stop and explain why.
Finally, give me:
The top 3 investments to scale
The 3 initiatives that need intervention
Any overlapping or redundant AI investments
Where we should redirect budget
The 3 questions leadership should answer before the next investment review
Don't assume missing data. Clearly label assumptions and distinguish between measured results and projected benefits. Prioritize business impact over model sophistication.Get more such prompts in the Prompting Playbook (free for you)
Stay curious, {{first_name | reader}}
PS. If you missed yesterday’s issue, you can find it here.
