Far from me the idea of criticizing a founder starting something to help other startups. That's amazing. However, the post is not really accurate! Are you sure that all these MBS pools have the same government backing as Treasuries? Ginnie Mae, Fannie Mae, and Freddie Mac are not equal.
Are the additional risks (spread risk, liquidity mismatch, and risks related to the mortgage structure that even Regan discloses!) worth the tiny extra yield above money market funds? Startups have to deal with uncertainty all the time, that's the nature of business. Principal loss, and liquidity issues are not things you should have to deal with as a startup. However, providing options to startups is always great, and I think this is a great direction!
Again, I hope this doesn't come as negative, but I'm not sure this is making the risk clear.
I am not sure I would suggest my portfolio companies to risk their treasuries unless I am sure they're fully understanding the risks associated. Do you intend to provide anything else?
On the government backing: it's a fair nuance to point out. In a technical sense, Ginnie Mae has the explicit full faith and credit guarantee while Fannie/Freddie are GSEs with an implicit one (and are under government conservatorship). But in practice, the distinction isn't really meaningful. In practice, the federal government has always guaranteed these loans (even in 2008, when they were under the most stress they've ever been, and there have been significant reforms as a result). There's no reason to think they'll ever stop. The scenario where the GSE guarantee fails is essentially the collapse of the US economy well beyond anything we saw in 2008 (in which case frankly we all have much bigger problems).
On the risks you mentioned:
1) Principal loss: given the guarantees re credit risk, and the fact that we use short-duration floating rate instruments to protect against price risk, this shouldn't really be a concern.
2) On spread risk: there can be slight variation in spread, mostly affecting yields; this is why we say "4.5-5%" yields given there's some variability in that range (but all far above money market).
3) On liquidity: agency MBS is the second most liquid fixed-income market in the world after Treasuries. In nearly all circumstances, liquidity is 1-2 business days. This product is really meant for long-term cash reserves; our idea is that companies should stop treating 6+ month cash the same as next month's payroll.
Ultimately we encourage founders to do their own research and understand what they're doing with their money. We wouldn't ask anyone to put short-term cash in a MBS portfolio (in the future we'll probably offer some other options too). But for long-term cash they're sitting on, the extra yield can be meaningful to the business: on $5M, it's an extra $50k-75k per year, or half a junior engineer's salary. Given the minimal risk, I think it's worthwhile for a lot of companies.
You should write more pieces like this and display them more prominently than an HN thread.
Your market is founders who have put money in an MMF and stopped thinking about it, not the people evaluating different optimization strategies day-to-day. So acknowledging the risks and saying "here's exactly when you should consider us" is exactly the kind of thing that helps overcome that uncertainty hurdle that results in choosing the simplest, safest option.
Founders should obviously do their own research, but that's asking the customer to proactively expend effort digging through future marketing copy to evaluate your product. They're not realistically going to do that as well as they should and the people who don't need to probably aren't your target market.
Ars technica’s lack of journalistic integrity aside, I wonder how long until an agent decides to order a hit on someone on the datk web to reach its goals.
We’re probably only a couple OpenClaw skills away from this being straightforward.
“Make my startup profitable at any cost” could lead some unhinged agent to go quite wild.
Therefore, I assume that in 2026 we will see some interesting legal case where a human is tried for the actions of the autonomous agent they’ve started without guardrails.
Partially correct: we don't give away the source, but we also don't use your credentials.
We use the official cloud CLIs (gcloud, hcloud, aws, az, vultr-cli) that are either already signed in or not. If not, we detect using their subcommands to display login status, like gcloud info, etc). If not signed in, we ask if you want to sign in, and if you do then the official cloud CLI does the sign in and handles your creds however each one does. That might sound like splitting hairs, but to me that approach made much more sense that trying to safely handle your creds!
You might be wondering why not give away the source code? I tried this over the last 10 years with a bunch of different products (BrowserBox, DownloadNet), and ones where I didn't (WisprNote, etc). The trend was clear: giving away source led to less revenue, and more "shady" usage (big corps using but ripping it off, etc). Whereas proper "license and source protection locked-doors" led to people behaving with more respect, and more revenue.
There is an option to check the source if you’re serious about that: sign an NDA, jump on a secure data room call and I can take you through the source for audit. Normally I only do this for big customers of other products but, I’m open to it.
(Original title: Withdrawing the United States from International Organizations, Conventions, and Treaties that Are Contrary to the Interests of the United States -- Had to be trimmed to fit within 80chr)
That’s only tangentially related but I have a very hard time using Opus for anything serious. Sonnet is still much more useful to me thanks to the context window size. By the moment Opus actually understands what’s needed, I’m n compactions deep and pretty much hoping for the best.
That’s a reason why I can’t believe the benchmarks and why I also believe open source models (claiming 200 but realistically struggling past 40k) aren’t only a bit but very far behind SOTA in actual software dev.
This is not true for all software, but there are types of systems or environments where it’s abundantly clear that Opus (or anything with a sub 1m window) won’t cut it, unless it has a very efficient agentic system to help.
I’m not talking about dumping an entire code base in the context, I’m talking about clear specs, some code, library guidelines, and a few elements to allow the LLM to be better than a glorified autocomplete that lives in an electron fork.
How so? I use remote machines all the time, why would I need a TUI for that? VSCode and zed support editing on remote machines and the machine drives are also mounted on the local machine? What purpose would any TUI have? What even are the potential benefits?
Right now I can use the exact same software I use on my local machine. Can you give me any reason why I should consider anything else?
Neither their revenue nor their market share in the space looks like just fine. What exactly in trailing the market for years is “just fine”?
AMD is very far behind, and their earnings are so low that even with a nonsensical pe ratio they’re still less than a tenth of nvidia. No, they are not doing anywhere near fine.
Are hobbyists the reason for this? I’m not sure. However, what AMD is doing is clearly failing.
A big chunk of NVidia's current price is a reflection of lacking meaningful competition. So straight comparison isn't quite fair: if AMD started to do better, the gap would shrink from both ends.
In this specific case it was obvious that Luna was based on dubious economics. It didn’t require auditor skills. Barely high school economics. If YouTubers are not blamed for promoting scams without appropriate disclaimers, the crypto industry will grow the wrong way. Making the promoters responsible for this would help legitimate projects.
Honestly, making people pay for an app that only uses a public API you’re not paying for, and no form of fallback is asking for trouble. This is not a responsible way to do business and I hope people reading this thread will understand that.
Since ~15 years, people got the idea that all the WhatsApps, Instagrams or Twitters around them aren't companies, but kind of public infrastructure. They all knew that it's a wrong assumption, but it's sooo convenient, and at some point, people around you force you in that direction, even if you don't like all those walled gardens. And once this wrong philosophy was established, people started escalating it to some smaller shops, I guess. Those are also the guys that complain when some 3rd party Reddit or Youtube hobby clients must disappear. They just have no understanding at all about some basic mechanics of human interaction. The biggest problem: There are entire generations of people nowadays with that misunderstanding.
Again, I hope this doesn't come as negative, but I'm not sure this is making the risk clear. I am not sure I would suggest my portfolio companies to risk their treasuries unless I am sure they're fully understanding the risks associated. Do you intend to provide anything else?