captainkrtek

Depends on the nature of the breach, if its a flaw in the software itself, then perhaps exploitable both in the hosted offering and BYOC offering, though perhaps the BYOC aspect gives a bit of "security-through-obscurity" and less discoverable as a target.

I get its imperfect, but I pay for Adobe Acrobat to work on legal documents (infrequent, but frequent enough that I have to pay), and PdfCraft works well enough for me that I cancelled my subscription.

very cool, fun to visualize :-)

It's frustrating to find AI so prevalent in writing. It takes some nugget of a human written idea, dilutes it and increases word count, only to make the actual core idea less obvious (or ultimately summarized by someone using AI).

Amen to this.

I've bounced back and forth on my feelings for AI and have landed in the realm of: - there are certain things it is exceptional at that humans cannot replicate. - there are certain things I do not want to use it for.

And review falls squarely in that first category. Similarly, it is exceptional at working through "low hanging fruit" type problems such as spotting inefficiencies, analyzing a profile to find flaws in software, etc.

Do any of the major streaming platforms have a stance against AI generated music?

I use Claude Opus 4.6 as an enterprise user, and have also noticed a lobotomization. In recent weeks it's been much more self-correcting even within singular responses ("This is the problem - no wait, we already proved it can't be this - but actually ...") I'm wary of 4.7 being a change in this pattern, it's frustrating to have such a substantial change in experience every few months.

On my team I've been adding additional linters and analyzers (some I've written with Claude) to run at CI or locally to prevent codified "bad patterns" from entering our systems. This has been nice as a backstop, as I can't enforce what everyone's Claude prompts and local workflows are, but we can agree what CI checks run before merging. Not a 100% solution, but it has been helpful so far.

I've been toying with this too.

I added a Claude skill (/gather-history) that consolidates the history of our session(s) specific to the change into a series of: decision log, involvement (how much did I write vs. AI, how many refinement iterations, reviews, etc.) that I can then include in the PR. So far this has been helpful for my colleagues to understand how I arrived at the change and how thoroughly it's been developed.

There seems to be so much value in planning, but in my organization, there is no artifact of the plan aside from the code produced and whatever PR description of the change summary exists. It makes it incredibly difficult to assess the change in isolation of its' plan/process.

The idea that Claude/Cursor are the new high level programming language for us to work in introduces the problem that we're not actually committing code in this "natural language", we're committing the "compiled" output of our prompting. Which leaves us reviewing the "compiled code" without seeing the inputs (eg: the plan, prompt history, rules, etc.)

One challenge with code review as an antidote to poor quality gen-AI code, is that we largely see only the code itself, not the process or inputs.

In the pre-gen-AI days, if an engineer put up a PR, it implied (somewhat) they wrote their code, reviewed it implicitly as they wrote it, and made choices (ie: why is this the best approach).

If Claude is just the new high level programming language, in terms of prompting in natural language, the challenge is that we're not reviewing the natural language, we're reviewing the machine code without knowing what the inputs were. I'm not sure of a solution to this, but something along the lines of knowing the history of the prompting that ultimately led to the PR, the time/tokens involved, etc. may inform the "quality" or "effort" spent in producing the PR. A one-shotted feature vs. a multi-iteration feature may produce the same lines of code and general shape, but one is likely to be higher "quality" in terms of minimal defects.

Along the same lines, when I review some gen-AI produced PR, it feels like I'm reading assembly and having to reverse how we got here. It may be code that runs and is perfectly fine, but I can't tell what the higher level inputs were that produced it, and if they were sufficient.

Will there be an interest in vision based wearables?

Google Glasses - dead

Apple Vision Pro - dead

FB/Meta x RayBan - dead soon(?)

It seems they can’t get over the social hurdle of having a camera strapped to your face, and the effects of that on people around you. I think the tech is neat, but not socially accepted as a concept to make it viable. My sister is big into tiktok and filming all the time, and it personally makes me hesitant to be nearby as I’m not comfortable being filmed all the time.