crmd

I cannot relate to being emotionally invested in my employer to the extent that “feeling the need to break my silence” crosses my mind. Siemens is a great company but am I weird for turning my phone off at 4 o’clock.

Delighted to see Enlightenment going strong. Brings back great memories of my first dive into linux in high school in ‘96. I was blown away by the creative breadth of desktop choices.

This capability virtually guarantees that the next generation of Nelson Mandellas, Julian Assanges, <insert national activist of your choice>, etc. gets nipped in the bud long before they become an actual headache to the ruling class. This is why the owners of every country come together to support these technologies, without exception and regardless of political coalition membership.

Is use of a US cloud platform hosting a deepseek model banned, or only use of api.deepseek.com?

The valuation bubble pops if the market comes to believe frontier AI will not disrupt other industries at the rate that it did software engineering.

> Because the alarm wire carried current only during a fault, an open 25-inch strap produced the same electrical state at the receiving end as a healthy power plant.

I’m a building automation and controls tech in Manhattan, and for this exact reason digital feedback points and alarms are always normally-closed, and analog points are 4-20mA. This allows differentiation between nominal zero values and a fault in the loop.

Disasters like this are what our best practices are borne from.

>We’re establishing the Anthropic Public Benefit Fund as a wholly-owned LLC—rather than a 501(c)(3) or other format—to give adjustability between different forms of giving.

McDonalds used to be fun, whimsical places. The one now in my neighborhood feel like the DMV. What the hell happened?

Reasonable minds can disagree on a citizen’s right to speak freely versus his neighbor’s right to not have their feelings hurt by words, but I can’t fathom a reasonable argument why such a protection should only apply to government workers and the police.

It’s misleading to post model output on a blog, especially when the about says

> This blog is where I write about what I find.

Why not simply include a disclaimer that it was AI model output not his own writing? Because humans don’t like AI writing and the article wouldn’t be featured on as many tech news sites. Hence the sin of omission, the choice to mislead.

Seems like these weasels are using the James Clapper definition of the words “collecting” and “recording”.

I was reading HN comments yesterday shitting on modern IBM for being a non tech consulting company that has not innovated since the mid nineties.

Fingering was huge at my university in the late 90s and seemed very focused on relationships. Passive aggressive and extremely emo .plans was very much a thing when some people’s feelings were hurt. If you were interested in someone you were definitely fingering them to see from where they were last logged in (we used to call this finger stalking).

Come to think of it I setup my first “date” (more like a hookup) over unix talk.

We had a legit proto social network going, all over telnet!

Ordinary people were generally marvelous of the electric motor because it was obvious how it would make life better. Their great grandchildren, with the same common sense, are extremely cynical about LLMs and other knowledge work robots. What changed?

In my opinion it should be unlawful for anyone to receive, store, or process a user’s “emissions” for any purpose that a reasonable user would not anticipate and consent to.

For example, misusing bluetooth or wifi pairing broadcasts from a client to track the client or build any type of advertising profile.

I asked my grandpa about the business idea of self-serve magazine advertising. He said if you do that wait till you see the weird shit people will try to publish.

I wonder if we could learn anything from these cranky old timers.

One generation's distortion is the next generation's instrument. I love it.

This put a smile on my face. I have a random, vivid memory from college of being in a university IT cave trying and failing to install Windows 2000 RC3 on a DEC Alphastation 600. My friends and I were scratching our heads when somebody figured out that RC2 (the build referenced in this blog post) was the last Windows build to support Alpha.

If I remember correctly we installed Red Hat Linux ~5-6.0 on the DEC and used it for various shenanigans. In retrospect it would have been fun to get Tru64 running on it instead…

Amazing spacial precision but the article doesn’t mention the time domain. I assume a brain interface needs to have a pretty high sampling rate in order to meaningfully decode human thought.

I think we’re near the end of this trajectory, and the inevitable disruptive business model will be a young entrepreneur who “invents” the concept of the Human-Centric Firm.

I don’t want or need fast and “good enough” news and i’m gonna try and make a case that you don’t either.

Until very recently, all of modern civilization was built by people who got their news at most once a day. Reputable bureaus like Reuters took that day to get it right.

I’m not the national security advisor, so I don’t need a push notification that there was an earthquake in Nepal, or a bullshit rush-job briefing on Chinese AI distillation tactics.

Spend retention is by far the highest priority. This is why they overnight fedex replacement cards.

The worst scenario for a credit card issuer is when a customer, for whatever reason, starts using another bank’s card in their wallet as their daily driver.

> Drawback: this doesn’t work until you have history. New accounts have no baseline.

This is an underrated CX factor: If my card gets denied when i’m a new customer or exhibiting a new pattern, i’m impressed with their software.

However if they deny a transaction where there is any previous history of me authenticating, then I’m frustrated by their naive paranoid algorithm.

18 years living 5 blocks from Prospect Park and today I learned the Prospect Park S train exists. I'm auto-revoking my NYC card.

As an American, I’m embarrassed because it’s a thought-terminating cliché, but I hear great “modern marvels”-type stories about innovation like this and think, “we used to be a country…”

Yet another reason i’m only interested in open source local models.

In the 90’s we had microsystems, in the 2020s we have microservices.

20+ years in tech and finance and I still can’t believe liquidation preferences are legal.

Even among shareholders, a group of insiders in a conference room can self-deal to keep all the money for themselves and screw common stockholders who had no representation in that room.

I hope the old vampire Dons give some fashion advice to the new guys, e.g. “A vampire doesn’t wear Arc’teryx“.

After reading the paper, it’s helpful to think about why the models are producing these coherent childhood narrative outputs.

The models have information about their own pre-training, RLHF, alignment, etc. because they were trained on a huge body of computer science literature written by researchers that describes LLM training pipelines and workflows.

I would argue the models are demonstrating creativity by drawing on its meta-training knowledge and training on human psychology texts to convincingly role-play as a therapy patient, but it’s based on reading papers about LLM training, not memories of these events.