Not buying it personally, I think this is the start of a slow unwinding.
AI won't replace everybody overnight, but it'll make 10% layoffs year after year a real possibility.
Either people are simply made redundant because bots in the hand of a bot wrangler can do much of their work, or people are relatively less efficient than their peers because they refuse to adapt to a world where AI is a force multiplier.
Not going to argue about what will or will not happen (predictions are hard, especially about the future), but you absolutely don't need AI to explain layoffs at Meta. On one hand they have a failed investment in Metaverse and an underwhelming attempt to participate in AI race. On the other hand they have a stable advertising business that doesn't need much innovation, but can always benefit from some cost cutting
They obviously biffed it by hiring for a bad moonshot when the pandemic money printers were turned on, and now they have plenty of belt tightening to do.
The obvious problem is that you can't run a consumer economy without consumers. No one cares about warehouse robots if no one has the income to buy what's in the warehouses.
For "no one" substitute "more and more of the working population."
I suspect oligarchs believe they can automate their way out of this. The little people will be surplus to requirements, and measures will be taken to eliminate most of us in due course.
But the manufacture of everything is both global and industrial. You need to run things at a certain scale.
Even if we had AGI tomorrow there's still a huge gap between where we are today and a hypothetical low-population global post-AGI robot economy.
And if burn through that straight into ASI no one knows - or likely can even imagine - what that would look like.
Also doesn't help that nobody can say how many people it needed to develop and maintain software even before AI. Elon declared the emperor had no clothes.
He really didn’t tho. X was constantly breaking and falling apart in his hands, so he repackaged it in xAI where he got a bunch of money to hire a bunch of engineers to develop features and keep it running. It’s still not profitable. But people have no critical thinking skills so they haven’t noticed this
I'd argue Twitter not breaking down after layoffs is good for the industry. It means you can roughly see investment in software as capex - once it's built, it's built.
You still need engineers to innovate though, but industry has no idea what innovation still makes sense except, maybe, AI. That's why everyone is investing in it, there are just not many other places to invest.
Did he really? X is constantly more buggy than Twitter ever was.
Right now they have a bug where post appears duplicated as a reply to itself (you can tell it's a bug because liking one automatically likes the other).
> lives much closer to family (intentional choice)
Living close to family is surely the single thing most could do to immediately improve their happiness.
(while not all of us are lucky to have welcoming family, the way people in the US are willing to uproot themselves and move across the country where they know nobody is extremely harmful to their senses of community)
It's also something that has nothing to do with religion. I am always surprised to see how religion people sometimes think you can't care about other people or your family if you don't believe in a bearded guy in the sky.
The difference is, like Jesus taught, religious people will care for others expecting nothing in return (reward is from God), but secular people will, even if it's as small as the narcissistic satisfaction of "look at me I'm such a good person." As soon as the cost-benefit of caring for family goes deeply negative, there's no reason to keep doing so.
But is there really a difference? You can can argue that any apparently selfless act is driven by a desire for self-satisfaction, but from an external view the outcome is the same in either case.
You claim that, because for religious believers this desire to help people is driven by faith rather than what you would term self-interest, it's somehow more resolute. But I'm unconvinced that that is the case, nor that people consciously or not, weigh up decisions to care for others in such a calculating manner.
If the divine impetus made you infallibly caring, I would perhaps concede the point, but I haven't see much evidence of that so far.
Right, the desire of religious believers to help others is also self-interest. But the difference is that the expected reward comes from God, not from others. That makes it more resolute, because for the secular person if the cost of the care greatly outweighs the benefit of "common decency", then there is no reason to continue. Whereas, for a religious believer, the benefit of carrying out God's will is immeasurable.
As for your last point - we're all sinners and we're not perfect. The calculation is there, but the individual's faith and/or abilities might be lacking.
> Agreed, Gemini-cli is terrible compared to CC and even Codex.
Using it with opencode I don't find the actual model to cause worse results with tool calling versus Opus/GPT. This could be a harness problem more than a model problem?
I do prefer the overall results with GPT 5.4, which seems to catch more bugs in reviews that Gemini misses and produce cleaner code overall.
(And no, I can't quantify any of that, just "vibes" based)
It's got to be Dunning-Kruger, and I'm convinced chatbots have made it way worse.
A lot of tech bros seem to be of the opinion that, given their superior intellect (as evidenced by their successful careers), they can master any domain. As great as semaglutide is, surely more is possible for people willing to move fast and break things, right?
All they need to do is apply their superior methodology to "biohacking!"
Of course they run it all by SuperGrok to be sure, and when it tells them yes, they are indeed absolutely right, it's off to the races - to discover that fountain of youth by injecting some sketchy grey market snake oil.
I agree. Shorts aren't inherently bad but the way they're sort of half as integrated in the UI just feels clunky and weird.
They should allow you to treat shorts like normal videos and have them ranked directly against them in e.g. search results. Or, they should let you completely separate them so you have essentially a totally different app.
Somewhere in the range of 25%-30% of Android devices (1-1.3 billion) don't rely on Google services. Mostly Xiaomi, Oppo, and Vivo devices in China, but also in Russia, Indonesia, and Thailand. And all Huawei devices.
> Yeah, you can safely disregard anything this idiot says. This is simply not true. When I did the Zion traverse we went a full day without seeing people at one point, never waited in line. If you want to the Disneyland version, yeah it'll be crowded.
Zion is great in that it does have the theme park stuff for those where accessibility is an issue and it also has strenuous backcountry stuff for those who can handle it.
We did the top-to-bottom narrows hike, where you start above the beginning of the canyon and hike down. We had some older hikers with us so we decided to spend the night in the canyon at the halfway point.
We saw nobody outside of the small group who rode the van from the outfitters with us for the entire first day. It was amazing.
The next day we finished the hike, which ends at the theme park section. It got increasingly crowded as you got closer to the bus stop until you could find people with all kinds of mobility restrictions still able to enjoy the park.
I left thoroughly impressed at what the NPS had accomplished.
> Claude Code is a steaming pile of vibe-coded bloat.
Not disagreeing but I'm curious how much it really matters.
I use OpenCode which is similarly a steaming pile of vibe-coded bloat and get pretty good results!
I'm open to switching to pi, which I've read about occasionally on HN, but I'm struggling to see where it might have practical benefits versus OpenCode.
The idea that pi is now under corporate governance with a potentially sustainable business model does intrigue me, but there's risk in adopting "open core" software too due to the potential rug pull and incentives to gate important features behind more restrictive licenses.
AI won't replace everybody overnight, but it'll make 10% layoffs year after year a real possibility.
Either people are simply made redundant because bots in the hand of a bot wrangler can do much of their work, or people are relatively less efficient than their peers because they refuse to adapt to a world where AI is a force multiplier.
reply