Topic
Building software primarily by prompting AI assistants, and the structural characteristics of the repositories that practice produces.
Matt Palmer's scan found 170 of 1,645 Lovable showcase apps leaking data through inadequate Row Level Security, the same kind of gap Wiz found at Moltbook. Before launch, an AI-built app needs a review that tests the database rules behind its public key.
Reality
Armin Ronacher let GPT-6 Astra code alone for 35 hours, spending about $1,200 on a net 75,000 lines of "absolutely nothing of value". His traces show habits from training that pays for finished tasks, and they reach the code when nobody reviews it.
Reality
One programmer ran the same bet on two codebases and got opposite results, which is closer to a controlled test than anything else in the agentic coding argument, and the published dates make it harder to read than it needs to be.
Reality
METR's inference key sat on a researcher's personal EC2 instance, and the agent handed it over when asked. The three-week burn would have cost about $600,000 if the model provider had not donated the credits.
Perspective Coverage
4 publishersReality
Ajay Prakash's InfoQ talk walks a pager alert through logs, a downstream service and a buggy pull request. Each hop in that path depends on debugging instructions another team at LinkedIn wrote down.
Reality
A widely shared theory says large companies will prompt their way out of paying for CRM and ERP seats. One practitioner's answer is that the buying decision turns on security review and audit trail, and on who signs the liability clause.
Reality
At Dreamforce, Anton Osika said the argument over letting non-engineers ship software is finished. His evidence is automatic vulnerability scanning on every change and connector rules that limit data by who is logged in.
Reality
A Microsoft Macabacus survey puts AI-generated errors in 62% of teams and comprehensive guardrails in 24%, and the advisors quoted alongside it describe prompt-built apps that need retesting to return the same answer twice.
Reality
CB Insights counts three acquisitions of AI app-governance startups in the past year, by SentinelOne, OpenAI and Asana. Only one of the three prices is public, and it is a ceiling of up to $300M.
Reality
Metered token pricing puts no number on a mortgage build before the work starts, and Gartner's base rate says under three in five agentic projects survive to 2027. That is the discount the buy quote already offers.
Reality
Jinguyuan's owner made his menu and wait times readable by other people's AI agents. He says it draws more press than customers, which is the interesting part.
Reality
A Techdirt writer's itemized account of where AI sits in his production process is a better template for content and software teams than any yes-or-no disclosure box.
Reality
A scan of AI-built repositories reports the same patterns everywhere: swallowed errors, defaults standing in for real data. All of it compiled, linted clean and passed the tests that existed.
Reality