Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
zmmmmm
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
17 ms
·
421.
▲
by
zmmmmm
1y ago
it's not really the number of characters so much as the cognitive load of having to do something different here vs there and anything I run successfully on the command line can't be directly lifted over into scripts etc. Along wit
422.
▲
by
zmmmmm
1y ago
that makes me feel much better!
423.
▲
by
zmmmmm
1y ago
> Instead of > > source .venv/bin/activate > python myscript.py > > you can just do > > > uv run myscript > This is by far the biggest turn off for me. The w
424.
▲
by
zmmmmm
1y ago
no I'm talking about the general concept of having ChatGPT passively able to read sensitive data / browser session state. Apart from the ever present risk they suck your data in for training, the threat of prompt injection or mode
425.
▲
by
zmmmmm
1y ago
Being "anti-web" is the least of its problems. This thing is an absolute security nightmare. The concept of opening up the full context of your authenticated sessions in your email, financial, healthcare or other web sites to Chat
426.
▲
by
zmmmmm
1y ago
Oh yes I'm completely with you on wishing Github would support PlantUML. It's almost trivial implementation wise, it's highly embeddable. It's precisely because I do like it that I want it to scale up better to more comp
427.
▲
by
zmmmmm
1y ago
It's definitely much needed. I use PlantUML for most diagramming but for anything with more than about 5 components in it I'm spending 20-30% of my time desperately trying to tweak the layout with hints. It's an interesting a
428.
▲
by
zmmmmm
1y ago
I think it's typically said to be about half as powerful as the M2 was. So far far less than the M5. I'm honestly quite surprised at what they seem to be able to get out of it on the Galaxy XR. I guess the direct partnership with
429.
▲
by
zmmmmm
1y ago
> I always felt this was such an outrageous burden to developers but the point is that it's not a burden? You get it for free. Unless you mean having accommodate in your app the fact that someone else's might be "shading&q
430.
▲
by
zmmmmm
1y ago
Sadly the whole open ecosystem is in grave doubt moving forward since Google announced they are shutting down side loading :-(
431.
▲
by
zmmmmm
1y ago
Sadly the weakest part of it is the old Qualcomm chip in it. It can just manage to render the 4k displays but it's barely capable of it even at 72 hz.
432.
▲
by
zmmmmm
1y ago
Cardboard was what got me interested in the whole area so it had it's value. But I think for about 90% of people it had negative value because it presented such a poor experience to the general user. To this day I ask people if they&#x
433.
▲
by
zmmmmm
1y ago
It's possible for sure that Google will abandon this and I absolutely recommend anybody considering buying in do so only on the value they can see and realise immediately, not on any future promise. But none of that takes away from the
434.
▲
by
zmmmmm
1y ago
Kind of sad to see here on "hacker news" that 80% of the comments are low effort cheap shots. The interesting thing here is the core of it, being Android XR and its deep AI integration, especially the spatial awareness. Devices wi
435.
▲
by
zmmmmm
1y ago
that's really interesting - thanks!
436.
▲
by
zmmmmm
1y ago
I'm curious how much trained in bias damages in-context performance. It's one thing to rely explicitly on the training data - then you are truly screwed and there isn't much to be done about it - in some sense, the model isn&
437.
▲
by
zmmmmm
1y ago
It's good to see experts with similar scepticism about agents that I have. I don't doubt they will be useful in some settings, but they lean into all the current weak points of large language models and make them worse. Security,
438.
▲
by
zmmmmm
1y ago
> If I invest in a homelab that can host something like Qwen3 I'll recoup my costs in about 20 months without having to rely on Anthropic For me it's equally that I don't trust any of these service providers to keep mainta
439.
▲
by
zmmmmm
1y ago
Yep, the ecosystem is well on its way to collapsing under its own weight. You have to remember, every system or platform has a total complexity budget that effectively sits at the limit of what a broad spectrum of people can effectively inc
440.
▲
by
zmmmmm
1y ago
if anything, installing the app spoon fed to you by your phone OS provider should get the pejorative. Let's calling, "Lameloading" or something to really nail it home.
441.
▲
by
zmmmmm
1y ago
It's a bit disturbing for the open model ecosystem, that your model could arrive with one of the elements of the lethal trifecta already compromised. I guess it was always possible any model could have adverse behaviour trained into it
442.
▲
by
zmmmmm
1y ago
I guess you could ask, why offices look like they do? they could already be imaginative spaces but companies mostly opt for "bland" spaces. Probably because if you're going to combine the preferences of 30 people in one space
443.
▲
by
zmmmmm
1y ago
I like that theory although it is depressingly grim ... the top dog species will inherently see any alternative intelligence as a threat and eradicate it. Would definitely make one pause for thought about the wisdom of creating an AGI ...
444.
▲
by
zmmmmm
1y ago
AGI is so last year. Now it's all ASI which is great because it was achieved in like 1968 or something so nobody trying to achieve it can possibly fail
445.
▲
by
zmmmmm
1y ago
yeah i don't disagree about breakfast specifically if it's completely standalone. However it's worth recognising that autistic people in burnout lose a LOT of executive function. Something seemingly simple like making breakfa
446.
▲
by
zmmmmm
1y ago
it's absolutely what all people do. The missing piece is you need to (a) multiply the severity/effort required by an order of magnitude and (b) add in a whole host of co-occuring factors like anxiety, OCD etc. that often come alon
447.
▲
by
zmmmmm
1y ago
The interesting thing about AI to me is it's got a bit of a "meta" effect in that it's giving generic software engineering a boost by increasing productivity there. So you have direct application of AI replacing existing
448.
▲
by
zmmmmm
1y ago
This is a really interesting aspect of masking. The thing is, it's often not the direct presence of external scrutiny that drives the feeling of needing to mask, it's a lifetime of having normal behaviour expectations reinforced t
449.
▲
by
zmmmmm
1y ago
part of the challenge with autism is that nearly all it's features are shared partially or individually by ordinary people as pretty routine character traits (eg: socialising is energy draining for "regular" introverts). It&#
450.
▲
by
zmmmmm
1y ago
sure - that's the data center model and nVidia will probably always own that. But I disagree that only privacy conscious enthusiasts will want to run locally in the end. Right now in the hype froth and bubble and while SOTA advances fa
More ›