vb-8448

15 ms·

271.
by vb-8448 1y ago
oh man ... it's so wrong on so many levels ... first of all it's not true, law enforcement are already doing their work today without chat control. In case we have massive chat control, do you really expect criminals to send plain
272.
by vb-8448 1y ago
> as a society we do need some sort of targeted backdoor into communications why you are assuming this is true?
273.
by vb-8448 1y ago
the real question is: when and in what form will it be re-proposed next?
274.
by vb-8448 1y ago
Actually nothing new here, this was the same in the pre-cloud era where everyone in enterprises prefer big names(ibm, microsoft, oracle, ecc) to pass the responsibility to them in case of failures ... aka "nobody get fired because of b
275.
by vb-8448 1y ago
> Maybe I'm just the old man screaming at cloud (no pun intended) but when did people forget how to run a baremetal server ? It's a way to "commoditize" engineers. You can run on premise or mixed infra better and chea
276.
by vb-8448 1y ago
gemini used to be the top for me until gpt-5 (web dev with html/js/css + python) ... and also with gpt-5 around it's doing its job, but it's really slow.
277.
by vb-8448 1y ago
maybe for very complex use cases, but I'm using htmx(and unpoly) + alpinejs + localstorage and still didn't find a case that doesn't fit.
278.
by vb-8448 1y ago
just out of curiosity: did you consider unpoly.js or alpine-ajax instead of htmx? If yes, why did you choose htmx instead of others?
279.
by vb-8448 1y ago
claims against gpt-5 are huge! I used to use cc, but I switched to codex (and it was much better) ... no I guess I have to switch batch to CC, at least to test it
280.
by vb-8448 1y ago
It depends, if you have to retrain existing models using new data(basically updating the cutoff date) it will cost you much less than today ... and I'm pretty sure today they are aiming for speed and "intelligence improvement&quot
281.
by vb-8448 1y ago
According to the data in the post, the cost of running the model for open AI in 2024 was 2B and if we strip out all training/research costs they had a loss of about 1B, peanuts. They can raise a little bit the standard subscription pri
282.
by vb-8448 1y ago
Obviously not, but my point is that they aren't losing money because it's intrinsically non-profitable on mass scale, but because of the market in this specific period.
283.
by vb-8448 1y ago
Nice read, but I'd add an objection here: even if models don't improve any more, and they raise the standard subscription to 100$/month, I'd still buy it (and a lot of other people, I guess) because I'd extract far
284.
by vb-8448 1y ago
the sens of my comment was: I would have liked that k8s looks like swarm.
285.
by vb-8448 1y ago
i'm the only one that wished docker swarm became the standard instead of k8s?
286.
by vb-8448 1y ago
In general, I think any human generated content in pre-2022 is valuable because someone did some kind of validation (think about stack overflow answer with user confirming that a specific answer fixed their problem). If they start to feed t
287.
by vb-8448 1y ago
I'm used to doing the same task 4 or 5 times (different sessions, similar prompts), and most of the time the result is useless or completely wrong. Sometimes I go back and pick the first result, other time none of them, other time a mi
288.
by vb-8448 1y ago
So, I guess they run out of data to train on ... I wonder on how much they can rely on the data and what kind of "knowledge" they can extract. I never give feedback and most time (let's say 5 out of 6) the result cc produce i
289.
by vb-8448 1y ago
There are still plenty of IBM-I (formerly as400) and AIX installations around Europe.
290.
by vb-8448 1y ago
So basicaly grok
291.
by vb-8448 1y ago
I guess the point is that to be a "real business" it needs to be at some point profitable.
292.
by vb-8448 1y ago
And token will burn and provider will bill XD
293.
by vb-8448 1y ago
> LLMs are not search engines, and I'm not gaining any followers or customers in any meaningful way because an LLM indexes my site. ^^^^ This For the moment, and for the foreseeable future, you are just giving your content for free
294.
by vb-8448 1y ago
They will produce PR(and probably shitty code) on a rate you are not able to review XD
295.
by vb-8448 1y ago
It's safer have them completely isolated, but it's slower and more expensive. Sometimes I just realize that CC going nuts and stop it before it goes too far (and consume too much). With this async setup, you may come after a coupl
296.
by vb-8448 1y ago
Hi, thank you for your feedback. I think in the past, I've already had a look at your repo, but it's a little bit different from my idea. Your is build around BPMN (and, if I'm not wrong, it's basically a easy way to cre
297.
by vb-8448 1y ago
A workflow engine, but not an "airflow like" (even if I hope it will become an airflow killer) but more similar to old job scheduling systems from mainframe era(OPC/TWS/CA7 ecc). In my opinion, those mainframe-era dinosa
298.
by vb-8448 1y ago
I didn't see it in the article, can some tell me what is the scale of " many writers."?
299.
by vb-8448 1y ago
I think the issue is that JS force you to have hundreds, or even thousands, dependencies. Python and other language have a richer std library and more "general purpose" packages, so the total number of dependencies is lower.
300.
by vb-8448 1y ago
I wonder if at this point it really matters who used whose data ...

More ›