Posts tagged with "spacexai"

2026-09-17 / tags: big tech, guides, meta, privacy, social media, spacexai

Just here to find a privacy frontend? Jump to the list of services.

Ever heard the phrase "have your cake and eat it too"? Social media companies love doing this. Thoroughly enshittified platforms like Reddit, YouTube, and Instagram gathered goodwill by offering a fun product to the public for free, then they leveraged the critical mass of people they attracted to shove ads down their throats once they were too locked-in^[1] with network effects to want to leave. Any time you browse a social media site, or especially a mobile app installed on your phone, these companies are collecting thousands upon thousands of valuable data points to construct a profile of who you are that they can sell to advertisers, whether or not you're logged in to an account.

Never let it be said that I begrudge people their choice of digital junk food. Today I will show you how to doomscroll without giving away your data.

Enter privacy frontends

Privacy frontends are a unique adversarial tool in the fight against big tech. They are web applications that access centralized social media services on behalf of users who want to avoid trackers and bloatware as they watch cat videos or peruse dank memes. In addition to avoiding data collection, users benefit from lighter resource usage and, much of the time, fewer technical issues. The downsides are that this tech goes directly against social media companies' business model, so they have a big incentive to go after them with whatever tactics they can, whether it's changing their code to be more difficult to work around or filing legal challenges, as X is repeatedly attempting with the Nitter project, to varying degrees of success (shout out to the Nitter project for refusing to take these attacks lying down).

Because of the adversarial nature of these projects, they are hosted in a decentralized manner in multiple instances. In computer^[2], an "instance" is jargon for an item that does the same thing as other items like it, as defined by its "parent", which it inherits from. In practice, this means that you'll need to find a list of instances for a particular privacy frontend you want to use, and pick from them depending on which is currently accessible, and any other criteria you'd like, such as where the server is located. Then, just navigate to that instance's URL in your web browser, and go bananas. A well maintained privacy frontend will have a list of instances available as part of its code repository, with a pointer to that list in the README (front page).

From there, you can use the frontend as a web application, starting from the home page and going where you'd like; or, you can find a link to the original social media site, copy the path (everything after the top-level domain, usually .com, including the first slash), and paste it at the end of the URL of the frontend's homepage.

Just as a quick demonstration, let's say I got a link from my buddy of an unfortunate typo someone made as they were getting mercilessly slaughtered online:

https://www.reddit.com/r/2007scape/comments/zdo61b/killed_this_noob_at_chaos_elemental_today_pleae/

I definitely want to see what's going on there, but I don't want to give Reddit my birthday, social security, and first born child for the privilege. What's a girl to do? I'll have a list of projects for various services at the end of this post, but we can start with a privacy frontend for Reddit called Redlib. That page shows its README, which has a link to their instance list. I'm in the US, so I'll pick one of the US based ones, https://redlib.catsarch.com. Now I'll just copy the path from my original link (remember, this is everything after the top-level domain, usually .com, including the first slash) and paste it at the end of this one to get:

https://redlib.catsarch.com/r/2007scape/comments/zdo61b/killed_this_noob_at_chaos_elemental_today_pleae/

Now I can view this extremely embarrassing moment of some random gamer's career, enshrined in history, without giving up my identity. Pretty neat! Note also that many of these services have instances accessible via Tor hidden services, I2P, and Yggdrasil for extra layers of protection.

List of services

I'll try to keep this section reasonably up to date with new info, but I can't promise that it will be current at all times. That being said, if you know of something that currently works for a major social media site, is actively maintained, and isn't listed here (or if something listed here is broken), feel free to drop me a line!

• Gelbooru: Gelbooru-Go (instances)
• Reddit: Redlib (instances)
• Imgur: Rimgo (instances)
• Instagram: Kittygram (instances)
• Tenor: Mezzo (instances)
• YouTube: Invidious (instances)

2026-09-05 / tags: ai, anthropic, google, openai, quick reads, spacexai, technical

Two days ago, big LLM-as-a-service providers had issues. OpenAI, Anthropic, SpaceXAI^[1], and Google^[2] all had issues at the same time.

It was later revealed that SpaceXAI's Memphis datacenter experienced an outage (X.com link). This knocked out Grok as well as caused issues for providers that lease compute from SpaceXAI, namely Anthropic and Google. Notably, OpenAI also uses Google infrastructure, though it's unclear if their outage was because compute they get from Google includes some of what Google gets from SpaceXAI. Another possibility: as providers hit the floor, end users and harnesses^[3] switched to their next preference, immediately shifting demand to their next provider of choice and overloading it.

The result? Much of the demand for LLMs-as-a-service that these companies have spent years cultivating had nowhere to go for agonizing hours, laying bare a fundamental – and fundamentally dangerous – assumption that organizations and individuals reliant on them have frequently made: that these services can be depended on for consistent availability.

Even if you set aside the incredibly unstable and shaky financial foundations of these services, which do pose them existential risk, the fact of the matter is that these APIs are not exactly renowned for their reliability – or cost stability. Years ago I read an article from an entrepreneur who had built a business on the back of someone else's business, and that other business decided to switch things up somehow, destroying the former. Whether it's pulling free APIs, shutting down entirely, or integrating your product themselves, other parties make moves you might not like all the time. A sustainable business is not overly reliant on external sources for data, services, or other needed resources that can't be gotten elsewhere.

If enough AI datacenters irreversibly stopped working tomorrow, you or I might be fine, but a gargantuan number of software shops would be incapacitated. As coding agents like OpenAI Codex and Claude Code have proliferated and made writing software quicker and less repetitive, businesses have adopted them at breakneck pace to keep up with their customers' demands and stay ahead of the competition. Usually, the traditional emphasis on time-to-market does not make room for concerns like "OpenAI might jack their prices" or "communities are uniting against datacenters and digging in their heels in fights to stop construction".

Earlier in the year, I saw an interesting development where individual techies, with their decidedly shallower pockets compared to VC-funded startups, got ahead of the curve on sustainability: they bought out all Mac Mini stock to run an application called OpenClaw, an AI agent that was often given irresponsible amounts of access to resources by tantalized users. Now, I encourage being very careful about sandboxing and monitoring with agentic systems, and I don't think LLMs are a one-size-fits-all solution for "intelligence" problems, but there are good uses for the tech. If you're going to run an AI model, running it on hardware that you own and can pay the costs of maintenance, electricity, and networking for is probably the best way to do it, no matter the scale. This setup is less like stashing thousands of dollars under the mattress instead of a checking account, and more like assembling a desktop from parts yourself instead of buying a prebuilt machine: you get all the cost savings and skills, and when things go wrong, you have the agency (and, hopefully, the capability) to fix them yourself. You're never relying on a set of nearly entirely unregulated tech companies to fix your rig's Wi-Fi not working – no more two hour long calls to ASUS. Doesn't that sound wonderful?

The snag with assembling your own desktop is that it requires a level of technical acumen that not everyone has, but lots of people need to use desktop computers. Software engineers already have that acumen^[4] – in most cases, this move isn't gated behind learning a totally new set of skills. Hardware may cost money, as it always has, but Ollama is free. You know what isn't free? LLM APIs – and engineer salaries for the time those are down.

2026-06-25 / updated 2026-06-25 / tags: ai, anthropic, openai, quick reads, spacexai

Elon Musk (SpaceX), Sam Altman (OpenAI), Dario Amodei (Anthropic), and a whole bunch of their buddies in and around the billionaires club are in the final stages of pulling the rug. The AI bubble has worked like every other bubble has: convince a lot of people that something is worth way more than its actual value; rake in a lot of money promising unlimited future growth; and make sure someone else holds the Beanie Babies before everyone realizes they're not going to autonomously cure cancer or end civilization.

If your bubble gets big enough, and you have enough insiders in on the racket, then you get to [rewrite the rules to the Nasdaq index](rewrite the rules to the Nasdaq index) to force passive investors to buy those shares of a company from the egomaniacal grifter who brought us the wildly successful Hyperloop whose official stance on its path to profitability is "we are going to launch data centers into space".

Uh huh. Sure.

A lot of people have been suckered into believing that this technology is as revolutionary as the three men I named at the beginning of the post, and countless AI boosters on LinkedIn, are claiming. Well, sure, if you can get past the fact that the AI companies have been navigating diminishing returns for nearly two years already, they've started charging you to use the models based on how much they actually cost to run, and as a result everyone is flipping out about how much money they're lighting on fire. The simple fact is that the fundamentals are not there.

Speaking as someone who worked at a financial firm: Wall Street does not know better than you. They are being conned like every other C-suite exec that gets a call from an OpenAI or Anthropic rep, getting pitched a miraculous technology that replaces 90% of their workers (who they'd love to fire and keep all that sweet, sweet salary money for themselves), and is accurate enough to increase the business' productivity. Meanwhile, actual, real businesses that invest heavily in AI just end up wasting more and more of their own resources trying to justify the sunk cost, and on and on it goes.

People want to believe the hype and bury their heads in the sand. This has proven throughout history to be a wildly ineffective strategy for accomplishing anything other than getting run over. We must be clear eyed about what is coming, and prepare and protect ourselves to the extent possible, dominant narratives of "too big to fail" be damned. Right now, in my humble opinion, one important piece of that is being very careful and thorough about zeroing out exposure to SpaceX (and, when they IPO, OpenAI and Anthropic) stock, including via funds that hold shares in those companies, and minimizing exposure to the tech sector more broadly.

Large language model technology is useful. It is interesting. It is categorically not useful and interesting enough to warrant OpenAI, Anthropic, and SpaceX to be currently valued at what is likely upwards of $5 trillion combined^[1], a number I expect to show a further rise before a precipitous crash. We are all going to suffer when that basic truth can no longer be avoided. A sinking tide lowers all boats; make your plans accordingly.

Update 6/25/2026: I got feedback on this post that inspired me to add some clarification. I do not mean to say that anyone should sell all their stocks and hold cash, which is also called "timing the market" and "a bad idea". What I am saying is that it makes sense for folks to consider reallocating their equity holdings to funds that track a different index, like the S&P 500. In a retirement account, selling holdings is not a taxable event, but in a regular brokerage account it is, so keep that in mind as well. At the end of the day, I am not a financial advisor or fiduciary, just someone who can see the cracks crawling up the foundation. If you have someone to help you navigate your personal financial situation, absolutely trust them over me with those matters.