groby_b

44 ms·

841.
by groby_b 3y ago
Video Games industry survivor here, chuckling at the low number ;) Yes, people do. It carries potentially huge personal costs, but if people are passionate enough, they absolutely will risk that. (Mind, I'm not saying this is a good
842.
by groby_b 3y ago
Kind reminder that starting a software business still requires you to, you know, live and stuff :) Even if you can build it all by yourself, for most people runway is limited by savings. And that means that with a few exceptions, it's
843.
by groby_b 3y ago
Ah, apologies for getting that wrong, and thank you for correcting!
844.
by groby_b 3y ago
Yup. As to the "why" part, the whole thing operates under the legal assumption that if you already shared your data with somebody else (e.g. your bank) you have no reasonable expectation of privacy. Essentially, the "secure i
845.
by groby_b 3y ago
Ouch. German ATC software in the 90s seems to have been a special place. (My favorite war story is "Hey, we have somebody 10 miles out from FRA airport, low on fuel, ATC software is down, can we escalate" - yes, yes we can.)
846.
by groby_b 3y ago
What this website provides is (and I am sorry to say this to the folk that put hard work into making this happen) lowered friction to camping. And a large number of folks do not understand LNT (see also: Instagram spots) and will be extreme
847.
by groby_b 3y ago
If software is deduplicating images by SHA including metadata , it's missing a lot about photography workflows :) So, huge props (and I mean that - so many tools get that wrong!) on your approach of trying to identify asset groups and
848.
by groby_b 3y ago
This is all a very neat summary, but I'll go out on a limb and say that for most of us, the first-order approximation should be "don't. Just don't". Been there. Done/sort of doing that. Have impressive collecti
849.
by groby_b 3y ago
As a request, this is extremely underspecified :) Job titles in our industry are notoriously vague, so it'd help to hear what you consider "more senior engineer" (I've seen definitions range from "more than 2 years
850.
by groby_b 3y ago
> am told they're full because of too many new members lately. This is good. This is a group that works well.They likely can give advice on where to ask, and whom to avoid. "Aww, I would've loved to play with you. Sorry t
851.
by groby_b 3y ago
It's only of limited helpfulness because the charging system for US healthcare is utterly bonkers. Might well be the actual care runs you $3500, but then you need to add $1000 for the aspirin they gave you, $800 for some gauze, and ano
852.
by groby_b 3y ago
Guess what. Mexico has a public health care system. You know, one of those evil "free" ones.
853.
by groby_b 3y ago
Yeah, no. Improved tooling to address common problems is absolutely the future. "Self-healing" is a long way off, because LLMs lack reasoning abilities to determine if a fix is good. And "makes the test pass" isn't
854.
by groby_b 3y ago
Except there's no provenance or root of trust. There is (IIUC) no back-propagation of a penalty if sites violate trust, just an overall observational measure. And I'd still say pagerank did work really well in an Internet where th
855.
by groby_b 3y ago
> But where does that leave us as a society? I'll worry about that once we address what it does to us as a society that CEOs outearn workers by 400:1 while oddly almost never being affected by layoffs. Same goes for smaller business
856.
by groby_b 3y ago
Add a trust metric and chains of provenance. Bad ring link -> bad trust percolating up that chain. Little trust, your site isn't always shown as part of the ring. Too much loss of trust, you're out. (Ultimately, this is a bad f
857.
by groby_b 3y ago
> On the other hand people’s initial reaction to much modern art is “what the hell? you call this art?” There's a lot of people who look at React and think "What the hell? You call this useable?" as well. The point of spen
858.
by groby_b 3y ago
You took 8+ years to get an MFA?
859.
by groby_b 3y ago
Why would you think there's an ASIC/FPGA design that significantly improves over GPUs specifically targeted at running large models already? Where's the win? The fundamental limit for hardware acceleration are number of gates
860.
by groby_b 3y ago
Neither the fact that it completely hijacks the back button, nor the fact that it plays sounds without consent does anything to endear this to me. I'd suggest giving those areas a bit of polish :) As a practical question for how you le
861.
by groby_b 3y ago
If you want to make a claim that we won't even see change coming, maybe the large mammals->humanoids transition isn't the greatest example. That took several hundreds of thousands of years. They did, to some extent, see it comi
862.
by groby_b 3y ago
I think that depends on who did the predicting :) There was a lot of "ooh, it will catastrophically fail within weeks", which was fundamentally an assumption that the previous team was entirely incompetent. (Any halfway decent tea
863.
by groby_b 3y ago
Yup. Same reaction. No explanation + "Book a Demo" -> KTHXBYE
864.
by groby_b 3y ago
A time-limited right of use. If you're a good steward of the land, it gets renewed. Leaseholds are (or at least have been) a fairly common and workable thing.
865.
by groby_b 3y ago
By that logic, people who have an injury should just be euthanized? Obviously, you wouldn't do that. Neither should we let species die off deliberately because we currently assume climate change will threaten their habitat. We do not k
866.
by groby_b 3y ago
None of the architectures we know imply in any way giant leaps in intelligence. That's the main sticking point. Thinking more does not equate thinking better. Or even thinking well. As for "animals implement them", it'
867.
by groby_b 3y ago
Does it? Does it really? Yes, we've seen huge improvements. We're still not close to human intelligence level. (No, test taking doesn't count). Worse, we've got pretty clear evidence that the training material increase f
868.
by groby_b 3y ago
Any development trajectory that doesn't look very suspiciously like a logistics curve is a good starter. (Pretty much all LLMs I've seen show an exponential increase in resources for a less-than linear increase in capability latel
869.
by groby_b 3y ago
There is not a single shred of evidence we'll reach superintelligence any time soon. There's however a lot of evidence that regulation benefits incumbents. You can do the math from there.
870.
by groby_b 3y ago
I mean, I get that HN covers a wide range of interests, and I love that. But, let's be clear: this is a juvenile rant, at best. It's not an actual cafe. It doesn't propose anything actually workable. It shows no curiosity aro

More ›