port11

12 ms·

331.
by port11 5mo ago
I think you/we are in the minority. I’m surrounded by parents that start sentences with “ChatGPT told me…” or “I asked AI and…”. We’re often talking about something that the literature refutes, but the LLM was trained on a bunch of pub
332.
by port11 5mo ago
Dunno, I got it immediately, made me think of Mozilla.
333.
by port11 5mo ago
In their defence, tons of abuse happens and forces providers to check identity. I’m not saying there’s no other way, but any hosting company I’ve worked it has strict limits until you verify your identity.
334.
by port11 5mo ago
An hypothesis from my own consulting: firing an employee can be somewhat difficult, and their salary has quite some taxes on top. At my previous startup we’d rule-of-thumb calculate someone as costing us 140% of their pre-tax salary. Put th
335.
by port11 5mo ago
Token use? Non-deterministic outcomes?
336.
by port11 5mo ago
I can’t tell if this is sarcasm or if you have a Claude Max 10x subscription.
337.
by port11 5mo ago
The weather is weird everywhere. Portugal now gets weird semi-tropical storms. Belgium had nonsensical heat earlier this year and then hail rains, now we’re up for another heat wave in May. We’ve been warned about climate change, but I’m gl
338.
by port11 5mo ago
One can only dream of Gemini Spam-Detector 3.5, which will finally make YouTube useable again. I can’t believe the kind of crap that gets recommended in the Kids app.
339.
by port11 5mo ago
Yup. “Wait, let me circle back and fix this another way.” What they can never fix is that plenty of Pro users complain that they never get quota, models are always maxed out. I left, and I can’t believe how much time I wasted in AGY steerin
340.
by port11 5mo ago
The point is that other topics exist that deserve talking about. There is SO much talk about LLMs everywhere, and in this kind of event they will eclipse other, perhaps more interesting conversations.
341.
by port11 5mo ago
There’s so much complaining about their keyboard issues, and it’s really an infuriating part of the iOS experience. The phone being hard to grip/slippery doesn't help, no…
342.
by port11 5mo ago
I’ve used WebStorm for around 6 or 7 years, coming from Sublime Text. WebStorm was noticeably slower, but it did a lot of extra things. It still does, but it’s gotten really slow lately, locking for 3–4 seconds when scrolling a larger file.
343.
by port11 5mo ago
Large companies break down the work into small pieces, many teams, many layers of abstraction. As a developer, I actually enjoyed doing some customer support calls with users, to connect with how they use the software, and what’s not workin
344.
by port11 5mo ago
Part of the psychosis are AI usage mandates, where companies require a certain amount of LLM usage per worker. Of course these things are useful, but forcing them on workers is psychotic.
345.
by port11 5mo ago
At some point, you have to break legacy support. I do wish Wayland had been so compelling that everyone’d switch, but I doubt KDE has the resources to perpetually support both.
346.
by port11 5mo ago
The truth lies somewhere in the middle. Context: 20 years coding, 13-ish of which professional. Using LLMs for side projects, including a very big one. Also using them to help manage our home server. I’ve used 20-ish agents with OpenRouter,
347.
by port11 5mo ago
It’s still worth minimising how many companies get your data, and minimising the data itself. I’m not sure what data Apple and Google get specifically out of their car thingies, but it’s very easy to avoid using their car thingie.
348.
by port11 5mo ago
I don’t have a side on this particular fight. I’m using LLM-powered coding assistants to help me with a side project, and I enjoy DuckDuckGo’s AI assist thingie. All good. But I would prefer better filtering than “slap AI on it”. On the who
349.
by port11 5mo ago
I guess the pragmatic answer is that you don’t need AI for that. You need good filters. I don’t like Zalando one bit, but I’ll grant them that it’s easy to find the right clothes on their website because they have very good filters. LLMs do
350.
by port11 5mo ago
EDRi is in the minority, the EFF is in the minority, and so on. But someone has got to fight the corps, they can’t be the only ones dictating what’s socially acceptable.
351.
by port11 6mo ago
Twitter/X is a company, it calls itself whatever it wants, and is likely registered somewhere under a specific name. Okay, people don’t call it X, the same way people pronounce IKEA weirdly or refer to vacuums as hoovers. Language is a
352.
by port11 6mo ago
I suppose the model was trained in such a way, the smugness is a facsimile of reality. Other models offer concise, direct answers without these idiotic qualifiers.
353.
by port11 6mo ago
Oops, someone forgot languages evolve! Otherwise you must use Turkye, one of the Middle English spellings. The United Nations agreed to their request, it’s a minor thing to let people spell it the way that was requested. You don’t have to,
354.
by port11 6mo ago
On the Pro plan you can max 10 windows per week using Opus for planning? I’m impressed. Even with Serena and really tight context management, I use Sonnet for most planning and Haiku for implementation. That gets me through the week doing 1
355.
by port11 6mo ago
Gemini’s smug and over-confident “this is the gold standard in 2026” definitely leaves little space for nuance if you don’t know the subject matter. Human students would, hopefully, know they don’t know everything.
356.
by port11 6mo ago
I rarely use Opus for planning (in the Pro plan). Spec a feature in Sonnet, hand it to Haiku, come back for review. That’s a 5-hour window gone, sometimes 2. I hit my weekly limit around day 4, with 2 maxed out windows per day (and sometime
357.
by port11 6mo ago
I find the ethics of power generation, resource use, and pollution in a world struggling with climate change to be more of a challenge than whether a few people can run some software. And that’s coming from a Claude user that’s getting tire
358.
by port11 6mo ago
Sibling comment does a great job, but I just wanted to add that their Terms and Privacy Policy are simply not compatible with privacy-friendliness. I used to analyse PPs to detect usage of data brokers, and I’ll confidently say that these 2
359.
by port11 6mo ago
Could it be you’re using Chrome with the offline Docs extension? On Brave and without the extension, Docs isn’t nearly as fast for me — even if Proton remains slow. I also wish I could afford Proton as a non-pro user…
360.
by port11 6mo ago
We agree, thank you for the clarification. Have a nice day!

More ›