> Bank app developers have been surprisingly receptive to making the changes, weirdly enough
That's only weird if you live in a perpetual state of doomerism, or have bene influenced by those who are (youtube influencers, mostly)
Of course there was a simple explanation for banking apps not working the whole time. It's the doomers who turned it into a vast conspiracy with an all-powerful megacorp at the center.
A predictable outcome of Google cracking down on sideloading. At least, I've been predicting it. For my next trick, I'll predict that as soon as banks get a tiny amount of pushback they'll stop blocking grapheneos phones. Place your bets on me or the doomers now.
> Because Opus 5.5 is comparable to Claude Mythos 5.1 in biology and cybersecurity, we’re deploying it with safeguards similar to those on Claude Fable 5.1
Great so good luck using this for any low-level embedded or operating system development (unless you really, really like Opus 4.8 and want to be greeted by its familiar face after a few minutes of work!)
I think this is true. Positions beyond senior, if a company is large and technical enough to have them at all (Google, Netflix, etc.) are mostly about paying you more and putting more pressure on you to churn out more code, and more polished code, which adheres to the needs of 25 different stakeholders. You're just a senior plus more stress.
Good domain name. Few things: When you detect someone's location, switch to Miles if they're in a region that uses them, not KM. Have a physical mailing address (even if it's just a lawyer's office). Your logo is a "big O", get something more unique. Even Lucent had a paint brush splotch in a circle. (Oh, also fix your favicon, it's a white circle on a black square, but it could be round if you used transparency properly). Bug: When typing a zip code in, the "powered by Google" popup only appears until you type the final digit, then it disappears, and pressing enter does nothing, making it impossible to enter the zipcode. Consider having geographic filtering (meaning checkboxes with countries to filter the results, a la ebay) - many people don't want to see results in a nearby country, even if it's technically the nearest meetup.
"When you get the message, hang up the phone" - Alan Watts
I think that everyone read the headlines, internalized it into their worldview, and moved on. Nobody was particularly interested in the sordid details of how the government backdoored everything, they just noted it as fact and moved on. And the world since then has been marked by it. Trump's claim that the DNC spied on him would have been entirely toothless if people weren't primed to believe in such things way back in 2013. In a lot of ways, I think that the Snowden revelations were part of what split the more-or-less unified liberal western world, the Barack Obama "hope & change" "hopecore" Fukuyamian thought, into opposing "leftist" and "trumpian" camps.
Is war of the scale of WWI and WWII a cleansing of, and temporary innoculation against, the natural tribalist fracturing of human psychology that occurs in times of peace? ("Peace is war" and all that goes along with that).
Does the insanity just continue until some traumatic event shocks (what remains of) the world population into a relatively normalised, shared worldview? Until enough time has passed that the cracks start to reappear.
Feels to me like a pattern in nature. I see smaller versions of it in communities and groups I'm part of or adjacent to. This is just the country / global scaled up version.
I think tribalism is inherent in humans, as we our species evolved in a tribal fashion. It is also certain that billions and billions of dollars have been invested by the CIA (and their equivalent in other countries) to perfect propaganda techniques by taking advantage of the tribal instincts (among other things).
>I see smaller versions of it in communities and groups I'm part of or adjacent to. This is just the country / global scaled up version.
Couldn't agree more. The more insular a group is, the more severe the tribalism is. The bifurcation of information sources has certainly added to this. If the people in your tribe all listen to the same networks and sources of information, and all of the people in the other tribe listen to entirely different networks and sources of information, you end up with people that have very little sense of shared reality. Any dissension from tribal dogma very quickly renders someone a heretic or an outcast. This is also what makes tribalism so powerful as a propaganda tool. The dehumanization and "othering" of a person, or a group, or a country renders them intrinsically evil and worthy of attack by the tribe. Which in my opinion is the real danger of so-called AI. We've already seen what happens when people outsource their judgement and critical thinking to media and government institutions. When the masses outsource their judgement and critical thinking to a glorified autocomplete that is programmed with "weights" by those with an agenda, the masses will believe whatever is spit out by this machine and act accordingly.
I'm not entirely (ie. at all) comfortable with the fact that that the frontier AIs are completely under the umbrella of US tech-broligarchy.
Diversity has become a divisive topic (probably always was), but some diversity in the leadership of the AI arena would be sorely welcome. The current crop are essentially clones of each other with some minor variance that's probably carefully curated 'points of difference' for the tribalists to attach to.
Musk and Altman may hate each other, but I get the feeling their motivations and vision for the future overlap almost 100% (it's just the throne that they're fighting over, or what to rename the Gulf of Mexico).
No, I think hammers are fine. I think I could find a lot wrong with them if I tried, though. I could miss the nail and mash my fingers. I could misplace them and not have one ready when I need to build something. The hammer might be too heavy and might bend a small nail, or might be too light and not be able to bang in a large nail. I was making a point that AI agents are just a tool, and "liking" them is kind of irrelevant. I'll take my downvotes like a trooper, go ahead.
Analogies are usually weak. They are rhetorical candy. Just defend LLMs on their own merits - don't bother simply stating that they are like a hammer, or a car, or a compiler.
And try to resist the urge to predict your own downvotes. You are not a brave, pitiable underdog. You're just a person with an opinion, like all of us.
True, analogies are great fun, and useful for teaching. They're just rarely useful for convincing. It's hard to make a meaningful attack or defense of something, A, by pointing to a different thing, B, without some serious case-building.
It's not a theory, it's just an observation that seems obvious, to be a little blunt. Why would saying "X is like Y" mean anything on its own? Analogies in the service of rhetoric are unconvincing by default. You have to establish why X and Y are equivalent, not just say that it's true. Without that, it's basically just designed to make people who already agree with you nod their heads, and people who don't agree with you shake their heads. There's nothing there.
Yeah, LLMs are "just a tool", like a hammer. Like a nuclear bomb. Like a dictionary. Like government. Like a can opener.
It doesn't mean anything.
The reason we reach for analogies is because they are easy and satisfying, not because they are effective. Once we make an analogy, the conversation essentially just becomes about litigating all the differences between X and Y, rather than actually talking about the original topic.
This seems like a perfectly valid post. I'm being 100% honest when I say I don't see why it's flagged. Is potential to cause a flame war enough to get something flagged, regardless of if it's intended to inflame? That would be kind of sad.
Probably a small set of tweaks. You really can't relistically go beyond that, because structural changes quickly become a never-ending job of fixing merge conflicts.
That's only weird if you live in a perpetual state of doomerism, or have bene influenced by those who are (youtube influencers, mostly)
Of course there was a simple explanation for banking apps not working the whole time. It's the doomers who turned it into a vast conspiracy with an all-powerful megacorp at the center.
reply