I’d give an AI agent its own Mac for the same reason I’d give a new hire their own laptop. If I ask it to fix a bug in one repo, why the hell should it be able to open my inbox?
49
A free browser extension can have access to the work you’re paying to keep secure.
I’d spend review time on what a tool can access, even if it never shows up in the software budget.
We wrote up the patched uBlock Origin flaw:
sythelabs.com/blog/ublock-or…
1
87
Everyone tells founders to build something they love. I’m building Sythe Labs because of something I hated, compliance.
I spent 10 years as a software engineer. Every time I had to stop what I was building to sign a policy, dig up an old Slack conversation, or prove I’d done something six months ago, I wondered why the hell this was still my job.
A decade of thinking “there has to be a better way to do this” turns out to be a pretty good reason to start a company.
If you or someone you love hates compliance as well, let's chat.
sythelabs.com
2
76
The Railway hack made me rethink niche cloud hosting. I still like the speed and price. For client production, the deal looks different if I need the same host to get to my backups and bring the app back up. linkedin.com/feed/update/urn…
1
1
36
GitBook’s Nov 3 notice asks customers to rotate credentials for connected tools. I’d want each connection to have an owner who can find and revoke that access.
gitbook.com/status/security-…
46
If I test one AI model before shipping, I want evidence it’s the one serving customers. Otherwise, those results may not apply.
Anthropic’s FlashProve explores checking output against a committed model. Early research.
anthropic.com/research/prova…
2
25
Imagine an agent refactors a download endpoint, breaks a cross-customer access test, then changes the test to accept the download. CI passes.
I’d review that changed assertion as closely as the code. Which requirement justified allowing access that was previously denied?
48
My hypothesis: experienced builders are less likely to treat features as a moat because they know how quickly those features could be copied. If you're building SaaS with AI, consider why customers would stay once someone offers the same functionality.
1
53
I feel like most people are just saying anything at this point. Show me a benchmark that supports any of this.
Databricks CEO Ali Ghodsi says if frontier models stopped advancing, it wouldn't matter for most orgs because enterprise AI is bottlenecked on context, not intelligence:
"There's no agentic work that's automated the whole enterprise. That has just not happened. Why is that? The real reason is the models are smart enough, but they just don't have the context that exists inside of any organization."
"They have not been in every meeting. They don't know what's in everybody's heads. They don't know all the processes. There's always a couple of employees who know everything in every organization. You go tap on their shoulder, and everybody's like, 'Oh my God, what would happen if he or she quits?'"
"If you just fused that and gave that context into the AI models, just the frontier today, I think there's so much productivity gains you could get for any organization on the planet."
"For that, we actually don't need smarter models... If the frontier doesn't advance, it doesn't actually matter for the vast majority of organizations on the planet. They're just so far behind in the adoption curve of actually automating things and getting value out of this stuff."
@alighodsi
Read the latest Profiles of the Future by @bhorowitz and @mikemcg0 on Databricks & the Future of the Firm: a16z.news/p/databricks-and-t…
50
Salesforce is awful but this is a great idea. It's exactly the only way technical users will EVER use SF.
David Sacks (@DavidSacks) says Marc Benioff is letting Claude own Salesforce's front end, and risking disintermediation on purpose:
"He is integrating Salesforce into Claude. He's willing to allow Claude to be the front end, to have the sort of the primary user relationship."
"So in other words, Benioff here is willing to risk disintermediation in order to ride this AI wave because he knows it's what the customer ultimately wants."
"You don't want to go into Salesforce or any SaaS application to write your agents. You want to create agents in the AI program that you're using. And you want those agents to be able to run across all of your SaaS platforms."
"Why is this ultimately good for his business and not bad is because, like I said, I still think that you need systems of record. You need the AI agents to basically go to those systems, get the data from a canonical source of truth. You just don't want there to be any probabilistic variability around that."
"It's not a rip and replace."
"Benioff is saying that the AI is now the user interface. He's basically acknowledging that he's leaning into it."
- David Sacks, All-In co-host, on the All-In Podcast episode published August 29, three days after Salesforce and Anthropic announced their Claudeforce partnership.
1
2
62
WE'RE PROFITABLE IF YOU JUST LOOK IN THIS ONE REALLY SPECIFIC AREA OF OUR BUSINESS.
No one is profitable, everyone is over-leveraged, money is being passed around in a circle and called GDP.
Ignore investors, bootstrap a services business, make cash, retire and have kids.
If you're selling a $50/month SaaS tool, your buyer may already be paying $200 for Claude. They're going to try doing the work there first.
You're competing with a subscription they've already paid for, especially if they only use one part of your product.
46
An AI tool asking for my entire hard drive to edit one proposal is asking for too much.
Give it the project folder. Make it ask for anything else.
We already know how permissions should work. A chat box shouldn't change that.
50
I think QA Engineers will become the most valuable job in large scale software orgs. LLM generated code, LLM generated tests, LLM's "manually" testing apps.
True QA needs to be run by someone who understands the personas, the intent, previous versions, and direct user stories.
The amount of calls I've been on to "share notes" with other founders who cannot answer simple questions about their tech is staggering. Vibe coding is a net positive, in my opinion, BUT understanding agentic engineering macro decisions is a MUST.
2
1
336
Took my biometric data for me to make a new account only to get rug pulled on usage. Excellent!
Starting September 14, we're permanently raising standard weekly limits in Claude Code by 25% for Pro, Max, Team, and seat-based Enterprise plans. Until then, the current 50% increase will be in place.
Readers added context they thought people might want to know
While standard base limits are increasing by 25%, this change represents a net 17% reduction in weekly limits compared to the active 50% promotional limits users currently experience, as confirmed by the official Claude account: x.com/claudedevs/sta…
1
139




