[londons_explore]

> In 2024 global internet usage grew from 5.3 billion users to 5.5 Starlink grew by a similar absolute amount, from 2 million users to 4 million over the same time period,

Is this some AI answer or did you foobar this math by a factor of 100?

[runarberg]

Whoops, a standard off by a factor of 100 error.

StarLink got 2 million new subscribers in 2024. Meanwhile the internet got 200 million new users. So even if every new StarLink subscriber would be a new internet user (which is obviously not true) they would still only account for 1% of new internet users. The real number is off course much much much lower.

[chermi]

This is definitely a small number. But I don't think it tells the whole story. Not every n+1 is the same. New satellite hookups in rural places, especially poor rural areas, combat zones, emergency situations etc. are more impactful than a new wired hookup in a city where there's already wifi in the library, for example.

[runarberg]

You made the statement:

> It's interesting how if it's anti-elon, it's ok to complain about how the poor are causing the privileged some difficulties.

Now it is up to you to show that this has outsized influence on impoverished communities.

According to ITU[1] the number one factor for lack of internet access is economical. The price of internet access can be reduced with traditional infrastructure, but governments are often unable or unwilling to invest in the infrastructure needed to bring faster and cheaper internet connectivity to underserved areas. StarLink should in theory fit perfectly here, but in reality very few people from underserved communities, especially in impoverished areas, can afford StarLink, and keep being underserved. What makes this even worse is that in the rich countries (like the USA and Australia) underserved communities that had been promised infrastructure to bring the broadband internet are facing delays and cancellations because politicians believe the community can get StarLink instead (when in fact they cannot afford it). This is known as the Uber effect (from when politicians used Uber as an excuse to cancel public transit projects).

1: https://www.itu.int/itu-d/reports/statistics/2024/11/10/ff24...

[inemesitaffia]

If the community can't afford Starlink, surely they can run Fiber and buy equipment from Ciena and Nokia (Huawei outside the US) to light it up themselves since it's more affordable than Starlink in your view.

[runarberg]

That is not the point. These communities depend on their state for their infrastructure, the state was (hopefully) able to provide them with garbage disposal, sewage, electricity, running and clean drinking water. The state should also provide them with quality telecommunication infrastructure.

[inemesitaffia]

The state can give them Starlink.

eg Quebec, Nunavut, Wales, New Zealand, Zimbabwe.

Both actual users and Governments paying for it value it in ways you don't

[chermi]

You've mentioned the uber effect about 100 times now, I get it. I honestly don't know just how mad (I'd wager not very) the average city resident is that they have uber now instead of maybe some shitty transit system with two routes out of 20 promised in 15 years 500% over budget. I do VERY MUCH wish we built solid, modern, useful public transit on budget in a reasonable amount of time. But it's honestly wishful thinking at this point.(3)

Similar story for deploying broadband, especially last mile. The government hasn't been very good at that from everything I've seen (1, including starlink's failure).

As for starlinks deployments, I can't find good numbers, so perhaps I was overconfident. I wish I could find more examples, it seems like they could be doing much more than they are, but they are a for-profit company. Given that it can serve rural, poor, otherwise disconnected communities, would you be for or against using starlink to serve them through some government-backed/subsidized efforts?(2)

(1) ---The Rural Digital Opportunity Fund (RDOF) RDOF is one of the most recent and troubled examples. It was a $20.4 billion initiative to bring high-speed broadband to millions of unserved homes and businesses.

Massive Defaults and Questionable Winners: In the first phase, the FCC awarded $9.2 billion to over 300 companies. However, major problems quickly emerged.

LTD Broadband: The single largest winner, provisionally awarded $1.3 billion, was ultimately denied the funds by the FCC in 2022. The FCC determined that the company, a small fixed-wireless provider, failed to demonstrate it had the technical and financial capability to deliver the promised fiber-to-the-home service to nearly 600,000 locations.

Starlink (SpaceX): The fourth-largest winner, provisionally awarded $886 million, also had its award rejected by the FCC. The agency cited that the satellite technology was "still developing" and questioned its ability to meet the program's long-term speed and latency requirements.

Widespread Defaults: By 2023, bidders had defaulted on their commitments for over 23% of the locations they had won in the auction, leaving millions of Americans in limbo and forcing the FCC to try and reclaim those areas for future funding.

---The Broadband Technology Opportunities Program (BTOP) - 2009 This program, part of the 2009 Recovery Act, was a grant-based system rather than a reverse auction, but it provides a clear example of budget and delivery failures.

"The Road to Nowhere": One of the most infamous examples was in Eagle Butte, South Dakota. The Cheyenne River Sioux Tribe Telephone Authority was awarded a $36.7 million grant to build a fiber network. Years later, a report from the Department of Commerce's Inspector General found that after spending $28 million, the project had connected only 70 customers, was nowhere near completion, and was plagued by mismanagement.

(2) https://techafricanews.com/2025/06/18/starlink-proposes-us27...

https://southernafricantimes.com/rwanda-and-spacex-sign-agre...

"200 Terminals for Amazon Communities: In Brazil, Starlink has provided at least 200 terminals to schools and healthcare centers in remote Amazonian communities, providing a vital link for education and telehealth."

(3) Of course, I predict one possible retort that it's these very same "oligarchs" that are tearing down the government and rendering it unable to build public infrastructure, thus lining Musk's and friends' pockets. To that I'd say, look at very blue california's high speed rail. Yes, trump did just take away some their funding, but it's been like 20 years now? And they're about 300% over budget, and their timeline doesn't even including full completion, but we can safely say 15+ years behind schedule. I'm not blaming democrats, I'm saying there's a systemic problem with state capacity. It is, very sadly, just not reasonable to expect much from the government when it comes to building public infrastructure.

[runarberg]

I’ve seen you post about the “last mile” problem before on this thread, and I think you may be misunderstanding the problem and the solutions to it. The last mile is a problem in public transit where people who live in sparsely populated area and commute to higher density areas, where high quality or high capacity public transit opportunities exist for every leg of the journey except the last mile to their home (think Edgewood WA. to the Sounder train station in Puyallup). In case of transit the infrastructure exists for most of the way, but it is costly to provide transit options to connect with the infrastructure, the last mile in that case can be a rideshares, park-and-ride, ride hail, dockless scooters, etc.

In telecommunication the last mile is when a high capacity wire extends to a nearby (sub)urban center, but not to your house because you live 10 miles away from there (I know about the issue in the USA of having a wire across the street and have to pay ridiculous sums to connect it to your house; but that is a different issue). If you solve it by bypassing the wire altogether and opt for satellite instead, that is not really the last mile is it. A last mile would be to put up a 4G tower (or a few 5G towers).

An analogy in the transit space would be that because of the last mile issue, you opt instead to drive the whole way.

[chermi]

The last mile problem is a generic concept in any networked system with certain common, naturally arising topology. I do not misunderstand it in the least. You're right that that putting up towers would be a (recent) solution in many places. Satellite is also a solution that applies anywhere (maybe not the poles?) and which requires less built infrastructure, which, again, we're not good at (in the US). But given the assumption that I don't understand what I'm talking about, I don't see this conversation going anywhere productive so I'm stopping here.