_bypa

7 ms·

121.
by _bypa 9mo ago
Indeed: https://www.theguardian.com/world/live/2026/jan/03/caracas-e... Or in the words of DJT (quoted in article): > "We have the greatest oil companies in the world, the biggest, the great
122.
by _bypa 9mo ago
The US continuing a long tradition of interference in LatAm: https://www.batimes.com.ar/news/world/the-united-states-hist... edit: typo
123.
by _bypa 9mo ago
It's the penalty for being human: https://ianreppel.org/the-penalty-of-being-human/
124.
by _bypa 10mo ago
> engineers report the most mixed results on quality later in the survey (51% better but 21% worse, the highest “worse” of any role). Hardly surprising. PMs, designers, and founders do work that has no immediate feedback on right/wr
125.
Thing, Creature, or Mirror? The Standards We Set for AI(msthgn.com)
1 points by _bypa 10mo ago|0 comments
126.
by _bypa 1y ago
You'll be asked to add a 10% tip for the algorithm. And if you refuse you'll be down-prioritized for the next deliveries.
127.
by _bypa 1y ago
Feels like a discussion on game theory ought to have been included from the employee's perspective: 1. If you don't speak up, no one will ever know you had an idea or solution. 2. If you speak up and someone in power is offended,
128.
by _bypa 1y ago
Step 1: Hand over your email address. Step 2: Bye. Seriously, why would I give some random service my email address to send a cartoon of a banana? This feels more like a litmus test for basic online intelligence.
129.
by _bypa 1y ago
The post states the glaringly obvious. LLMs are statistical in nature like almost all ML algorithms. Expect deterministic behaviour from a non-deterministic model is to be surprised by what ought to be obvious.
130.
Opening the Door to Lifelong Curiosity(templeton.org)
2 points by _bypa 1y ago|0 comments
131.
by _bypa 1y ago
> We needed to nail the keyboard. We needed to make sure the text input works on this thing Then why, oh why, is the iOS keyboard still so atrociously bad and appears to be getting worse over time?
132.
by _bypa 1y ago
This is what Amazon has been doing for years with Marketplace: check what's popular and then compete on price.
133.
by _bypa 1y ago
I believe there are two reasons. First, there have been so many recent scandals, people are tired or unable to focus on yet another scandal, no matter how egregious. Second, it appears as if almost all the rich and/or powerful are invo
134.
by _bypa 1y ago
... in mice.
135.
by _bypa 1y ago
This. Most bosses are so obsessed with applying Pareto's 80/20 rule in all situations (even when it does not apply) that most would trade velocity for accuracy without thinking. Frankly, I doubt the average manager would know wron
136.
by _bypa 1y ago
> half-baked AI feature can damage core trust in your product. Except when everyone else does the same, because what are customers going to do: leave for a competitor that's just as bad?!
137.
by _bypa 1y ago
I guess no one at the Pentagon watched Terminator... and, no, I'm not worried about a sentient AI like in the movie. I'm much more worried about a) a buggy AI that confabulates the world into a conflict yet insists its right or b)
138.
by _bypa 1y ago
The Dawn of Everything by Graeber and Wengrow. It retells human history from the perspective of modern archaeology and anthropology with a focus on North American natives and their influence on European philosophy or what we believe it to b
139.
by _bypa 1y ago
The current best one- and two-gate errors are well below 0.01% and going down with every generation of chips. See: https://ianreppel.org/quantum.html There are no theoretical reasons QEC can't exist. In fact it already
140.
by _bypa 1y ago
Shameful yes, but entirely expected from corporations that only care about profits. Sorry, shareholder value.
141.
by _bypa 1y ago
In theory, yes. In practice, many never take responsibility for screw-ups and instead cash out before there are any consequences. When it comes corporate decisions, again yes. But most happen at much lower levels, where there is even less a
142.
Effects of smartphone restriction on cue-related neural activity(sciencedirect.com)
2 points by _bypa 1y ago|0 comments
143.
by _bypa 1y ago
No climate scientist ever claimed to be 100% certain.
144.
by _bypa 1y ago
You're running an inference on the server or sending requests to DeepSeek? If it's the former, mention it because that's good for data privacy. If it's the latter, all data ends up in DeepSeek's hands.
145.
by _bypa 1y ago
The image has illegible text and even says BapsyGPT.
146.
by _bypa 1y ago
Nice! You might want to add a description of what data is stored, as I see no mention of it. Don't expect people to copy-paste text without knowing what happens to it. And even if they do it's sensible to explain what you do with
147.
Researchers use machine learning to improve the taste of Belgian beers(physicsworld.com)
3 points by _bypa 1y ago|0 comments
148.
The Incompleteness of Ethics(aeon.co)
8 points by _bypa 1y ago|1 comments
149.
by _bypa 1y ago
One counterexample does not invalidate the general upward trend in global temperatures or the intensification of weather extremes.
150.
by _bypa 1y ago
$20 million may be the highest bounty offered so far, but I doubt anyone capable of such an attack would sell their idea for such a paltry sum to a vague entity.

More ›