Under US law, the NLRB excludes employers subject to the Railway Labor Act. xAI and X would need their own classification. It does not propagate automatically through ownership or mergers.
He talks about this new tech for extracting more value from engineers as if it were fracking. When they become impermeable you can now inject a mixed high pressure cocktail of AI to get their internal hydrocarbons flowing. It works but now he feels all pumped out. But the vampire metaphor is hopefully better in that blood replenishes if you don't take too much. A succubus may be an improved comparison, in that a creative seed is extracted and depleted, then refills over a refractory period.
Poets say science takes away from the beauty of the stars—mere globs of gas atoms. Nothing is 'mere.' I too can see the stars on a desert night, and feel them. But do I see less or more? The vastness of the heavens stretches my imagination—stuck on this carousel my little eye can catch one-million-year-old light. A vast pattern—of which I am a part—perhaps my stuff was belched from some forgotten star, as one is belching there. Or see them with the greater eye of Palomar, rushing all apart from some common starting point when they were perhaps all together. What is the pattern, or the meaning, or the why? It does not do harm to the mystery to know a little about it. For far more marvelous is the truth than any artists of the past imagined! Why do the poets of the present not speak of it? What men are poets who can speak of Jupiter if he were like a man, but if he is an immense spinning sphere of methane and ammonia must be silent?
A footnote for those of the millenial or more recent persuasion: we take the full “vastness of the heavens” as given, as we’ve seen it described pretty confidently all the way back to the science books of our childhood. But cosmology, and frankly the entire field of astrophysics, is strikingly young. The idea that nebulae are in fact whole independent collections of stars, and that the observable universe is large enough to accomodate all of that, is younger than quantum mechanics and relativity both, and only got acceptance after a huge fight. The name “Big Bang” was originally a pejorative used in a similar, later fight. And so on. When Feynman said this, the idea of nebulae as galaxies was younger (~40 years) than the key idea of quarks (confinement/asymptotic freedom) is today (~50 years), and I’m guessing the latter still counts as new and arcane in your mind.
I feel uncomfortable labelling nebulae as collection of stars. The more appropriate term is stellar nursery if you want to allude to their role in star formation.
They themselves are just clouds of gas and dust where protostars have begun to form.
Stellar clusters are what you would call a collection of stars.
Also on the note of cosmology and astrophysics being strikingly young fields, I think that's fair statement if we consider their modern definitions. Although their core ideas have already been discussed in a lot of ancient civilizations. It was a lot more philosophical and less rooted in science though (except for the observational astronomy, which remains perhaps one of the oldest scientific discipline).
Sorry, yes, there’s a terminological disconnect here: M31, say, is the “Andromeda Galaxy” to us, but the “Andromeda Nebula” to Hubble’s contemporaries circa 1920. The recognition that at least some of the cloudy (nebulous, literally) stuff in the sky is galaxies (and that the universe fits more than one) was the very point of the fight I mentioned. The world before it was thought to be drastically smaller in a way that I find difficult to think about.
That's right. Historically, nebula was anything that looked cloudy, so a lot of astrophysical objects that we now understand are distinct, were simply labelled as nebulous. M31, as you said, being a great example.
Modern astrophysics still carries the baggage of obsolete terminology to this day, from names of objects to names of units.
Once you start going down the rabbit hole you start asking questions like "does the photon oscillate?", "what exactly is resonant frequency?", "how different is the electron cloud around a molecule from that around its constituent atoms?", "how does a photon passing by/through a molecule cause its electron cloud to oscillate?" etc. The act of clarifying each to oneself in however simple a form is the insight we all crave. Good teachers like Feynman do a great job of it which is why their books are so highly valued.
PS: People might find the recent free book Atomic Physics for Everyone: An Introduction to Atomic Physics, Quantum Mechanics, and Precision Spectroscopy with No College-Level Prerequisites (2025) good for an initial understanding of atomic physics - https://news.ycombinator.com/item?id=46961595
I don't get the pressure. I don't know about you, but my job for a long time has been continually learning new systems. I don't get how so many of my peers fall into this head trip where they think they are gonna get left behind by what amounts to anticipated new features from some SaaS one day.
How do you both hold that the technology is so revolutionary because of its productive gains, but at the same time so esoteric that you better be ontop of everything all the time?
This stuff is all like a weird toy compared to other things I have taken the time to learn in my career, the sense of expertise people claim at all comes off to me like a guy who knows the Taco Bell secret menu, or the best set of coupons to use at Target. Its the opposite of intimidating!
> That said, handling contradictions more explicitly is something we’re thinking about.
That's a great idea. The inconsistencies in a given graph are just where attention is needed. Like an internal semantic diff. If you aim it at values it becomes a hypocrisy or moral complexity detector.
Interesting framing! We’ve mostly been thinking of inconsistencies as signals that something was missed by the system, but treating them as attention points makes sense and could actually help build trust.
As a corporate drone, keeping track of various internal contradictions in emails is the name of the game ( one that my former boss mastered, but in a very manual way ). In a very boring way, he was able to say: today you are saying X, on date Y you actually said Z.
His manual approach, which won't work if applied directly ( or more specifically, it will, but it would be unnecessarily labor intensive and on big enough set prohibitively so ), because it would require constant filtering re-evaluating all emails, can still be done though.
As for exact approach, its a slightly longer answer, because it is a mix of small things.
Since I try to track, which llm excel at which task ( and assign tasks based on those tracking scores ). It may seem irrelevant at first, but small things like: 'can it handle structured json' rubric will make a difference.
Then we get to the personas that process the request, and those may make a difference in a corporate environment. Again, as silly as its sounds, you want to effectively have a Dwight and Jim ( yes, it is an office reference ) looking at those ( more if you have a use case that requires more complex lens crafting ) as will both be looking for different things. Jim and Dwight would add their comments noting the sender, what they seem to try to do and issues they noted ( if any ).
Notes from Jim and Dwight for a given message is passed to a third persona, which will attempt to reconcile it noting discrepancies between Jim and Dwight and checking against other like notes.
...and so it goes.
As for flagging itself, that is a huge topic just by itself. That said, at least in its current iteration, I am not trying to do anything fancy. Right now, it is almost literally, if you see something contradictory ( X said Y then, X says Y now ), show it in a summary. It doesn't solve for multiple email accounts, personas or anything like that.
This was a really interesting read. Thanks for the detailed breakdown and the office references. The multi-persona approach is interesting, almost like a mixture of experts. The corporate email contradiction use case is not something we had in mind, but I can see how flagging those inconsistencies could be valuable!
The raw sync layer (Gmail, calendar, transcripts, etc.) is idempotent and file-based. Each thread, event, or transcript is stored as its own Markdown file keyed by the source ID, and we track sync state to avoid re-ingesting the same item. That layer is append-only and not deduplicated.
Entity consolidation happens in a separate graph-building step. An LLM processes batches of those raw files along with an index of existing entities (people, orgs, projects and their aliases). Instead of relying on string matching, the model decides whether a mention like “Sarah” maps to an existing “Sarah Chen” node or represents a new entity, and then either updates the existing note or creates a new one.
> the model decides whether a mention like “Sarah” maps to an existing “Sarah Chen” node or represents a new entity, and then either updates the existing note or creates a new one.
Thanks! How much context does the model get for the consolidation step? Just the immediate file? Related files? The existing knowledge graph? If the graph, does it need to be multi-pass?
The graph building agent processes the raw files (like emails) in a batch. It gets two things: a lightweight index of the entire knowledge graph, and the raw source files for the current batch being processed.
Before each batch, we rebuild an index of all existing entities (people, orgs, projects, topics) including aliases and key metadata. That index plus the batch’s raw content goes into the prompt. The agent also has tool access to read full notes or search for entity mentions in existing knowledge if it needs more detail than what’s in the index.
It’s effectively multi-pass: we process in batches and rebuild the index between batches, so later batches see entities created earlier. That keeps context manageable while still letting the graph converge over time.
There is a silver lining to this particular catastrophe. A large fraction of the abandoned tankers are a result of sanctions on Russia. It means they're working. Ukraine cares very much about these ships, having damaged a dozen tankers with their "kinetic sanctions" in the past year. They care because the oil is both a war material and a major source of Russian revenue. More abandoned tankers means a weaker Russia.
At the risk of sounding pedantic, you're (I think, implicitly) claiming that reducing Russian oil revenues doesn't impact their war capabilities. Why would that be the case?
I'd love to have a danger heat map displayed on a HUD while driving. Say a default green banner that goes red near a hot spot or even animates near a current hazard. Mostly it could use these same stats, but then be strident if anything unpredictable is detected nearby.
As I mentioned in another comment, Waze does this. There's a stretch of the Capital Beltway that, if it was on a race track and compressed, would be called "esses." It's totally fine to navigate at 80 MPH with no drama in any mechanically sound, post-1980 car, but it catches mediocre drivers by surprise. Waze throws up a "history of accidents" message whenever I drive through it.
Sustained winds in Dallas on Wednesday, Feb 4, were around 10–15 mph, with occasional gusts approaching ~30 mph. I wonder how well delivery drone station keeping works when the wind suddenly gusts by 20 mph.
reply