HN user
5.3 was incredibly better than 5.4/5.5. I stuck with it for months after 5.4 was released, and kept testing 5.4/5.5 every now and then but they both were too inconsistent, too rash. I switched to 5.5 a few weeks ago and now regret it, but I am no longer seeing 5.3 as an option to use, only 5.3-Spark, which is trash compared to 5.5.
When an architecture company seemingly uses AI to render mockups, they really need to ensure consistency and accuracy. It's not that difficult nowadays. It was quite confusing trying to understand the differences in design between pictures and to compute why the tunnel seems so short compared to the mountain, until I realized it must have been laziness; not laziness because they are using AI, but laziness to do their job right.
I was surprised to see Apple exposing the device boot time. Not sure why a typical app would need this, and in a world in which iPhones are infrequently rebooted, the combination of this with other fields must indeed be very deterministic.
Why is the author considering Claude Code a "real developer workflow"? Unless you're doing complex tool calling, is CC really resource-heavy?
If AGI were to happen, or if AI became a trillions-of-dollars-generating industry, you wouldn't want to have your data-centers which might be the most valuable thing on Earth be located in a foreign country. All this investment in infrastructure is not purely based on where the industry is now, but predicated by where those who are bullish about it think it will be in 5-10 years.
Here is a December look at the Astrovan II they used today: https://arstechnica.com/space/2025/12/nasa-rewraps-boeing-st...
Video from today: https://x.com/NASAArtemis/status/2039411225205886979
This is a highly sensational take that is basically fan fiction. From "the era of purposefully frustrating humans is over", to "the added bonus of the collapse of the US economy. Frankly, it’s well deserved." and "everyone in the world is rooting for the Chinese models"; nothing of that is grounded in reality.
The Chinese models are open source because they are not state of the art. Once they catch-up or lead, they will likely close them down by a government mandate. Just like Meta was fine with Llama being open source but once they started to get close to OpenAI/Google/Anthropic, they shifted their language to "maybe we won't keep doing that."
The idea that AI will end the "rent-seeking class" that has effectively existed for thousands of years is... not going to happen! The business model just adjusts. And if AI is going to be an economy-shaping super disruptor, the cloud-hosted models will continue evolving beyond what you could ever run at home under the desk.
I like this. It'd be great to see such a table of the key issues with proposed solutions, to highlight how the waste isn't an insurmountable impossibility to solve. Having said that, federal lobbying by the healthcare industry was $750 million in 2024 [1], and this is the blocker that needs to be addressed first to be able to enact change.
[1] https://www.managedhealthcareexecutive.com/view/health-syste...
If they are paying them 6-month severances like Block did, this means they are effectively saying 1,600 people for 6-months wouldn't have fixed JIRA's usability and performance, which if they could have done like many have been begging, they'd would probably make more money long-term than this firing would save.
The Neo is incredibly executed by Apple, and one they must have been planning for for years: to be able to create a machine this good at this price. I wouldn't be surprised if it dramatically reshapes the laptop industry.
Spec-wise, this is as good as an M1-M2 Air, which is already an over-powered device for most non-professionals. All the "compromises" they made, like no center stage in the camera, less ports, only one monitor support, "just WiFi 6E", and others, are all non-issues for a typical average consumers.
And the price is the best it could be. At $499 for students, in a year's time when Gen2 is released, you will find a new Gen1 at possibly $399, and a refurbished Gen1 at even less. I don't see why anyone who wants an "entry-level/starter laptop" would buy anything but a Neo. We already are in a world in which average people don't need specific Windows-only apps. Most common apps are either cross-platform or web-based.
Dell, HP and alike are lucky that they're being enamored with datacenter server demand. I expect them to shift-away from the consumer laptop market and focus more on the enterprise in the coming years, which could have negative consequences for their pro-lineups.
I love the website; the design, the video, the NSFW toggle, the simplicity.
I love the idea; definitely something I ran into a few times before and wish I had.
Unfortunately, I am not installing a closed-source daemon with access to the filesystem from an unknown (to me) developer. I will bookmark this and revisit in a few weeks and hope you had published the source. :)
Except that it doesn't need to be consumer to start off. You can build specialized robots that deliver value at a massive scale. Imagine a "Prep Cook" at a restaurant, there are millions of these around the world. If the Optimus can do that job for a price of $1,000/month, that's likely to be more efficient and better quality than a human can do. And there has to be many jobs like this.
Putting aside whether or not they are a fraud, their design looks so good, unlike Meta's ugly glasses. Which of course, might be because it's the one thing they've spent time on in the past 2 months of dev time, and they may not actually be accounting for any practical manufacturing realities.
The article mentions early a "cancer diagnosis" but puts that aside and moves on, when this is pretty much the crux of the issue. Prostate and Breast cancers are a 1 in 8 chance. The risk of no insurance at 25 is very different than 50, and than 75. And everyone at all ages is paying for those expensive treatments.
The system is broken, but going without insurance is you basically toying with the odds of life.
If CUDA isn't that strong of a moat/tie-in and Chinese tech companies can seemingly reasonably migrate to these chips, why hasn't AMD been able to compete more aggressively with nVidia on a US/global scale when they had a much longer head start?
A new M4 Air is now $799 at Amazon, and a new M1 Air is $599 at Walmart. So it's not like $999 is really the starting price if you spent a minute to search outside of Apple's Online Store.
This rounded corner change feels very off. Since Apple has that same radius across all its products (software and hardware), it could be signaling a broader upcoming shift in their hardware, perhaps driven by industrial design needs for future AR/VR/MR glasses.
Not OpenAI, but Anthropic CPO Mike Krieger said in response to a question of how much of Claude Code is written by Claude Code: "At this point, I would be shocked if it wasn't 95% plus. I'd have to ask Boris and the other tech leads on there."
[0] https://www.lennysnewsletter.com/p/anthropics-cpo-heres-what...
I don't really think Meta ever had a vision beyond "Facebook is a social network to connect people". Since then, their strategy has primarily been driven by their fear of being left behind, or of losing the next platform war. Instagram, Whatsapp, Threads, VR, AR, and now AI, they all weren't driven by a vision as much as it was their fear of someone else opening a door to a new market that renders them obsolete. They are good at executing and capturing the first wins, but not at innovating, redefining a market, or pushing the frontier forward; which is why they eventually get stuck, lose direction, and fall behind (Tiktok, Apple Vision Pro, AI).
Who will be responsible for figuring out what AI features to build? I think it is reasonable to look into it, seriously, with the point of view of "can we disrupt ourselves before being disrupted." This doesn't mean putting a significant engineering team behind this, but, to put a significant effort in figuring out what is it you could build, and what an ROI on that would be.
You've two outcomes from this, either you do find a disruptive AI angle and move a sufficiently-large part of your team to it, or you don't, but figure out a minimal effort that would satisfy the "investor positioning angle". The third option, to do nothing or aggressively push back against AI and the CEO's desire, would potentially yield to no Series C or a down-round, which is something that you, your CEO, and your customers would not like.
For each of the 2024 7 swing states, the winner was <1% ahead on average, so what good are these polls if the results are going to be within their margin of error?
They need to either find a more accurate way, or... give up!
I see this as a combination of three forces at play: AI, WFH, and Skillset--all adding downward pressure to hiring talent in the U.S.:
1) While A.I. may now be only adding 10-20% of productivity gains, the rapid pace of improvement leaves open the possibility that tha gains can be soon much more than that. So, instead of scaling your company now, if you can afford to, wait out a bit and see where this goes.
2) Even though much of BigTech is clawing back WFH, startups aren't as much. And once you introduce WFH to your culture and processes, it is hard to reason with the idea that you should pay $200K/year for an engineer when it can cost you a fraction (possibly 20-50% of that) to hire them remotely from another country, when also nowadays most of these remote employees are more than willing to work in EST/PST timezones. This used to be the case before COVID, but now many more startups have accepted and adapted to the idea of WFH.
3) While advanced skillsets and deep experience is necessary in many (but not most) startups, and while these skills are more difficult to find in India or Pakistan, the reality is, for many, many tech companies, most of the work doesn't require top-notch skills. You don't need a top 99% percentile in frontend engineering skills for a 1-year-old "name whatever category" app. And with the recent rise of focus on profitability, frugality, and the difficulty in fund-raising, being cognizant of cost per talent is now a thing.
I think Elon and Vivek's comments are more nuanced than they are taken. Elon, given he's at the cutting edge of engineering, must be having difficulty hiring top-99.9%-percentile talent against BigTech, and wants to open the pool of these types of talent from elsewhere. I don't think he wants H1Bs for React Native engineers. I am interpreting his comments as "I want to suck-in all A.I. researchers into America".
Well, there are quite a lot of rumors and stigma surrounding COBOL. This intrigued me to find out more about this language, which is best done with some sort of project, in my opinion. You heard right - I had no prior COBOL experience going into this.
I hope they'd write an article about any insights they gained. Like them, I hear of these rumors and stigma, and would be intrigued to learn what a new person to COBOL encountered while implementing this rather complex first project.
The title and discussion in the linked article describe this as "worse after firmware update", but the title here suggests Apple removed the feature entirely.
On my end, I've noticed I have recently been checking whether it was on or not. So, I am one more person experiencing some issue here.
Having been involved in such projects on the corporate level, I do agree that driving efficiency and reducing costs, even when everyone wants and tries to do them, is rather difficult, and very slow to implement. However, once you go down the track and commit to it, a major benefit is it raises the awareness of the importance of frugality and efficiency, and future projects do become more efficient. So maybe if you cannot make a huge dent like $2T, it could help slow down the increase in deficit.
Netflix Engineering is probably far ahead of any other competitor streaming service, but I wonder how much the ROI is on that effort and cost. As a user, I don’t see much difference in reliability between Netflix, Disney, HBO, Hulu, Peacock and Paramount+. They all error out every now and then. Maybe 5-10 years ago you needed to be much more sophisticated because of lower bandwidth and less mature tech. Ultimately, the only real difference that makes me go for one service over the other is the content.
I think Netflix for a few years now has been optimizing for "the cost of production of each consumed hour". They have been doubling down on acquiring/producing international content, reality tv and live specials—which are all cheaper, and less risky given the formula repetition. And when they have an expensive show that is not viral and super-successful, even if it was good by many standards, they cancel it. Many have noticed the reduction in quality and the increase in rate of cancellations, but it doesn't seem to be hurting their subscriptions much, at least not yet.
While it is true that most of the election-year policies are promises that end up not being fulfilled, the policies that are chosen by the candidate to push the most are also a reflection of their character and beliefs. And just like you've to distill the fake from real promises, you've to do the same for the character they project, because how they present themselves is not always any more real than the policies they promise.
Sam needs to hire a consultant when he wants to name products. First it was ChatGPT, then came the 4o's and o-1's, and now this. "World" is not easy to pronounce for some non-English speakers (the "rld" part). For a product literally meant to be used by everyone, that's quite the faux pas.
This video explains how they plan to catch the booster with Mechazilla [1]. The team at SpaceX has some serious guts to be doing this!