HN user
I don't think we are offloading thinking to AI. We just started to use it. AI is a tool useful to write text, for a fast searching, for the boring work. Personally I don't take the first answer, I like to challenge, to ask why, to tell it's wrong.
I wrote this story about the concept of community
the title is misleading, somebody could think that a startup launched a new steam based machine after the Hormuz closure
The BBC article doesn't specify the text with the image, but I clearly see a procedural gap in the police department. Accusing a man who only posted a photo, reorganizing the search based on an unverified photo, it's a big failure.
Did Orwell teach anything? What will they do with the next Visitors' spaceship photo?
Titles are really a trap cannot reveal the real work, everything it's related to the size of company, type of organization and sector.
A "top" position in a small company could be a mid one in a big tech but with a narrow field compared to the small one when sometimes you need to refill the coffee machine. At same time the flexibility of small one helps to solve problems of big ones.
An alternative, open and freely accessible OS for mobile computing is always good for a healthy market. Most of us have a limited view of the global market and don't know which areas prefer de-Googled OSs. If all of India or Africa decided to ditch Google, it would be a massive shift. We cannot forecast if the West will slowly decide to move to other solutions inspired by tech-savvy users or by becoming more privacy-conscious. It will take time, but desktop Linux is also slowly growing.
Governments that require age verification for operating systems to protect children also drop bombs on civilian neighborhoods, fight wars that orphan millions and tolerate child labor, exploitation, poverty.
History teaches us governments are the best at protecting children.
This law perfectly demonstrates the constraint problem: regulators assumed age verification is a simple checkbox at account setup.
Right now I'm on an ESP32 with free RTOS, will I need to add a keyboard and display just for age verification?
Registration just creates friction for legitimate developers (thousands) while bad actors simply rotate shell companies and fake/stolen IDs.
This conflates identity verification with criminal deterrence, they're not the same thing.
Look at Chinese smartphone brands: Xiaomi, OnePlus took their slice from Samsung.
Same pattern will play out with RAM and SSD. In 3-5 years, it won't be 'Samsung' on the label, it'll be 'Li Tech' or equivalent.
Western manufacturers ceded the market through strategic choice; Chinese companies are filling it systematically.
Western memory manufacturers decided to chase the AI bubble, abandoning the consumer and low-requirement markets entirely.
Chinese manufacturers are now capturing that entire segment with full vertical integration. When this bubble stabilizes, because it will (it's not going to grow to infinite), Western companies won't recapture those markets.
They've already ceded competitive advantage for the next decade. This is a structural shift, not a cyclical shortage.
It's another step in the transformation of Western industry that began in the '80s: the shift from real economy and human-centric production to financialized operations.
Not long ago, Italian Supreme Court (Cassazione, ord. 34217/2025) established a critical principle reflecting a broader trend in personal data management: the "digital image" of a person must reflect their actual current situation through complete and accurate information. The court rejected deletion but endorsed contextualization, articles must explicitly note the acquittal (assoluzione) or dismissal that concluded the investigation. An article mentioning "investigation for [crime]" without noting "subsequently acquitted" is factually incomplete, therefore false.
This principle applies directly to search engines and data processors. When Courtsdesk was deleted before reaching AI companies, the government recognized the constraint: AI systems can't apply this "completeness" standard. They preserve raw archives without contextual updates. They can't distinguish between "investigato" (under investigation) and "assolto" (aquitted). They can't enforce the court's requirement that information must be "correttamente aggiornata con la notizia dell'esito favorevole" (properly updated to reflect the favorable resolution of the proceedings).
The UK government prevented the structural violation the Cassazione just identified: permanent archival of incomplete truth. This isn't about suppressing information, it's about refusing to embed factually false contexts into systems designed for eternity.
I remember when I went from 286 to 486dx2, the difference was impressive, able to run a lot of graphical applications smoothly.
Ironically, now I'm using an ESP32-S3, 10x more powerful, just to run Iot devices.
I wonder if the bow drill principle for boring holes evolved from fire-starting techniques, where the same reciprocal motion was already understood and mastered in that years.
Just speculation, but it suggests how practical problem-solving builds on existing techniques rather than appearing fully formed.
I think archaeology requires a multidisciplinary approach that has only recently begun to emerge. For too long, especially in past centuries, archaeologists focused on history and languages while neglecting engineering, chemistry and the practical techniques that enabled survival and innovation. That's why the general public views our ancestors as 'primitive,' when in reality they possessed techniques many of which we've lost or still don't fully understand.
in spite of my age, I'm one of first digital native. I never used it and nobody explained me how to use. At same time I avoid most of analog instruments: multimeters, scope meters, calipers.
I agree that the ceiling of 4h is correct for high intense cognitive activities and that morning, after few hours after woke up, is the best time for these activities. And I would add that trying to do more the same day will affect the next day.
Meetings, phone calls are distractions, especially in the morning, they could also bring different thoughts far from current task.
To have the best result we must reorganize the company according to this, because most organizations prioritize visibility over results, but compensation, promotion and trust structures reward deep work instead of meeting attendance.
I've carefully read this interesting discussion: some political, mostly about full cloud services (AWS) vs partial EU providers, or lock-in vs indipendence.
I think the problem is elsewhere. The real advantage of big cloud players isn't their individual services. It's seamless integration and simplicity. We need a service integration standard for infrastructure that enables:
- Service discovery
- Networking
- Observability
- Configuration
This benefits everyone: EU companies, US startups, enterprises anywhere avoiding vendor lock-in. A standard letting services integrate regardless of who provides them.
Not just container orchestration (Kubernetes), but something working across bare metal, VPS, containers, and remote machines.
The irony is that EU education is still broader and more grounded in fundamentals, compared to US one that has become increasingly skills-oriented.
I also prefer to design solutions that are portable and platform independent, cloud providers simplify and hide something to you, it has a cost (not just money) that you cannot quantify on long term and that's clear for who has experience in both worlds.
If we review the history we can notice that there was always an influence from politics/religion to science, literature, arts, philosophy and the use of them by politics, maybe to justify some decision and state of facts.
It helps to empower control over population and fits perfectly in the social and historical context: the emperor blessed by God, the evolution theory, the epic poems, theory of race, the industrial revolution, and modern times don't escape these patterns too, we just suppose to be neutral.
I think that during ice age it was more easy to travel from Sicily to Malta, maybe they'll be able to find traces of humans even earlier than 8500 years ago.
Anyway I think we have a lot to learn from our ancestors, how were they able to move such heavy megaliths?
Security teams don’t fail by missing bugs. They fail by fixing the wrong ones.
Gartner’s EAP category shifts focus from CVE volume to real attack paths across cloud and identity. Most alerts never reach critical assets. EAPs show what actually matters.
I'm glad you find it interesting. I developed the theory at university, studying how ASK and FSK modems work. To build this, you’ll need to understand the Shannon-Hartley theorem, band-pass filtering, Fourier transforms, and convolution.
For the practical 'how-to,' I recommend studying GNU Radio and SDR++; they show how to process IQ data or raw audio streams directly, and for sure there are other libraries. On the 'ancestor' side, look at the AX.25 Packet Radio protocol and AFSK (Audio Frequency Shift Keying). These are the same 'softmodem' principles used in FidoNet nodes decades ago.
GSM Arena can help you find phones with integrated FM receivers. You'll notice that many features are market-dependent, meaning: the receiver is often physically present but simply disabled by software.
I never actually used Fidonet. I started on BBS systems just as the internet was becoming affordable, and I made the switch early.
However, I apply the concepts of FidoNet almost every day. I often design offline-first devices, where store-and-forward logic is a necessity, not an option. Many are deployed in remote areas where signals are never optimal, there a High-Gain Antenna is the only solution.
I also prioritize binary protocols over structured JSON; you have a much higher probability of delivering a few hundred bytes of binary data than a bloated text object when the link budget is tight. Finally, I never expect Wi-Fi to work beyond 5-10m when the router is placed inside the metal structure (that's why my skepticism about BT on cruise ship).
I don’t take concepts as a 'full package'. I evaluate what is worth taking based on the requirements. The brilliant part of FidoNet is the asynchronous persistence.
In a 'sea of supercomputers,' a real-time mesh (like Bluetooth) fails because it requires an end-to-end path right now. Store-and-Forward allows a node to hold a message until it 'sees' any valid peer, turning every 'meat-bot' into a mobile post office.
My main concern with this entire discussion is the reliance on Bluetooth to achieve the result.
If we truly want to build a free and open intercommunications system, we must put all ideas on the table, establish clear targets (a doomsday system or inviting a friend for a drink), and evaluate what is truly available versus what is not.
Only from that foundation can we begin to define a project that survives the real world.
I was pushing on the walkie-talkie case to gain the maximum results from existing phones, that's a true emergency case.
You’re absolutely right that the 5G/LTE baseband is a black-box nightmare to repurpose. But I’m not looking to hack the cellular modem; I’m looking for the dormant '4x4 car' already available.
For instance, many chipsets have an integrated FM receiver that is essentially a high-sensitivity VHF radio. By taking the raw audio output and applying a Software Modem (AFSK/FSK) in the user-space, you bypass the kernel/firmware complexity entirely. You don’t need to sideload a modem driver if you treat the audio jack or the internal FM bus as your physical layer.
The 'complexity' is real if you try to fight the manufacturer's fences, but it vanishes if you understand the full stack. A pair of wired headphones becomes your dipole antenna, and the phone's CPU becomes your DSP engine. It’s not about rebuilding the Ferrari; it’s about realizing there’s a VHF engine hidden in the chassis that doesn't need 'permission' to receive bits. You just need a software demodulator the catch them, but for sending you'll need an external transmitter (an USB SDR or jack-to-FM).
It depends on which step of the staircase, from pure hardware to pure software, you want to position yourself. Some projects require staying closer to the metal, while others can be purely software. I move up and down this staircase depending on the specific requirements.
If the requirement is to communicate where consumer standards like Bluetooth fail, like in a ship, you have to choose the system for the environment. I evaluate these choices like an architect building a robust system, rather than just using what is available at the nearby shop.
The smartphone is just an advanced walkie-talkie, currently limited only by the mobile operator, the law, the radio chipset, and the OS.
In a true emergency, who can stop you from modifying that architecture? Once you treat the device as an independent radio node (using its DSP power to run custom modems) you can establish a mesh network with a range of several kilometers.
We have a '4x4 car in our pockets; we’ve just been conditioned to treat it like a toy.
To have an FM receiver work on a phone, you do need an antenna, the wired headphones serve that purpose perfectly. An FM transmitter is easy to find; you can use the simple 'Jack-to-FM' adapters designed for car radios, or much better, a USB SDR (which can range from a few kHz to GHz).
Regarding the 'protocol for the waves,' you'll need to play with modulation. That’s the fun part. In technical literature, there are many well-defined modulations (like AFSK or FSK) with clear suggested applications for low-SNR environments.
As for Android support, I have no idea. I understand that in this thread, 'free' sounds like 'freedom,' but freedom has a cost. The freedom of communication requires investment: in hardware, software, and the time to learn the physics of the environment.
I’ve read all the posts and, as the 'old man of the village', I would suggest taking a look at FidoNet. It was running 40 years ago, for more than a decade, before the internet was available to the average person.
Store-and-forward, hierarchical organization, scheduled transmissions, working over dial-up and radio links, everything is there.
There is nothing new to invent, and it was far more reliable than the 10m real-world range of BT5 (not the 1km claimed for lab devices, which aren't commercial phones).
A BT5 mesh only works under well-defined conditions, which usually coincide with the cases where you don't actually need it.