fouc

10 ms·

151.
by fouc 9mo ago
I can't believe this was downvoted. It makes a lot of sense that it would be highly useful to have mass custom inference chips.
152.
by fouc 9mo ago
I disagree it's "toward the left" but I would also disagree if you said "toward the right". By that I mean I've observed BOTH extremes happening.
153.
by fouc 10mo ago
Haha yeah I didn't want to get into an exhaustive list
154.
by fouc 10mo ago
Suddenly occurs to me that it would be neat to pair a small LLM (3-7B) with an HN dataset
155.
by fouc 10mo ago
Oh huh, I forgot about that. I guess I downloaded X after the base system was installed.
156.
by fouc 10mo ago
I guess that means you don't live in the US, or in the UK, or in Australia.
157.
by fouc 10mo ago
He meant with X & web browser and so on. The QNX disk had gui + browser and a few other gui apps.
158.
by fouc 10mo ago
Photon microGUI was included in that, and it blew my mind that you could literally kill and restart Photon without disturbing any of the GUI apps that were still running. They also mailed a manual along with the demo disk, and I was amazed
159.
by fouc 10mo ago
Would be interesting to see how much a good "system"/server-side prompt could improve things. I noticed some animations kept the same sketch style even without specifying that in the prompt.
160.
by fouc 10mo ago
I think they were suggesting that it might be possible to inject the initial sketch into every image/frame such that the model will see it but not the end user. Like a form of steganography which might potentially improve the ability
161.
by fouc 10mo ago
Oh I was wondering why some of the hallucinations introduced Chinese text/visuals, I'm guessing that might be due to the negative prompt.
162.
by fouc 10mo ago
new utility command coming soon! wdtci - "what does this curl install?"
163.
by fouc 10mo ago
A lot of the responses seem to be blaming the user/learner and requiring them to change their mindset/attitude, which is actually an insane take. As you pointed out, SRS isn't the full solution. BTW, I would say that language
164.
by fouc 10mo ago
Pretty good, I've noticed the animation tends to veer off / hallucinate quite a lot near the end. It is clear that the model is not maintaining any awareness of the first image. I wonder if there's a way to keep the origina
165.
by fouc 10mo ago
That would require a browser that supports WebP
166.
by fouc 10mo ago
Does this work on a browser that doesn't support WebP? That would be useful.
167.
by fouc 10mo ago
The title for your submission should be a Show HN: Interesting point about the problem of HN's depth-first comment structure. It feels like perhaps HN should collapse sub-comments by default to encourage more breadth. Perhaps collaps
168.
by fouc 10mo ago
Oof, that felt too real. I'm half torn making that a reality before someone else does. That's often the thing about these torment nexuses, they're somehow profitable.
169.
by fouc 10mo ago
sometimes I find that it helps if my prompt directly names the tools that I want the LLM to use, i.e. I'll tell it "do a WebFetch of so and so" etc.
170.
by fouc 10mo ago
Nice one! I would've said half a microcentury would be the perfect limit for most school lectures, but for an engaging talk a full microcentury would be acceptable :-)
171.
by fouc 10mo ago
looks like a redditor snuck into HN
172.
by fouc 10mo ago
I feel like the solution is ultimately going to be some kind of trust-less or low-trust system that ultimately incentivizes every individual to do the right thing, no matter where they might be in the hierarchy. We can't rely on top-d
173.
by fouc 10mo ago
it's physically painful for everyone else to listen to you talking about your topic.. if you talk for too long that is. :)
174.
by fouc 10mo ago
Google Translate hasn't moved to LLM-style translation yet, unfortunately
175.
by fouc 10mo ago
>LETS GOOOO! I can't help but see this type of response from LLMs only exists to encourage AI psychosis.
176.
by fouc 11mo ago
Seems a bit sad/ironic that it sounds like the solution in OP's case would be to switch to Android for that exact behavior that your side didn't want. (And that switching to iPhone would bring that "feature" in) I
177.
by fouc 11mo ago
Interesting, wouldn't the MBP microphone be even further away than the HA's microphone?
178.
by fouc 11mo ago
> Cut off half of each lane at the merge point. Good point, even in a situation that necessitates a favored lane, could still setup this half-of-each-lane merge point well in front of that.
179.
by fouc 11mo ago
Sad. Yet another victim to the inexorable march of the enshittification of technology in the name of "updates" and "security".
180.
by fouc 11mo ago
I've long wished we had a standardized communication channel between cars. It could even be fixed status codes. I've always expected that in the future when all cars are fully self-driving, they would have some kind of communicat

More ›