I always laugh at these shots from the hip criticizing YouTube and Google. As though Google doesn't have a entire team of data scientists and top tier engineers managing this experiment and driving it to optimal results. (Spoiler: they do)
If you don't like the service, you can stop using it. And if you do, they have already factored that into their metrics guardrail, and it was the right decision.
>As though Google doesn't have a entire team of data scientists and top tier engineers managing this experiment and driving it to optimal results. (Spoiler: they do)
Optimal for who, though?
From Google's perspective I'm sure these changes push towards a more optimal revenue generation through ads. They potentially also push a more optimal layout on tablets/phones, or for shorts content.
Meanwhile from a desktop/laptop user perspective these changes are hardly optimal, especially compared to what they were before.
> If you don't like the service, you can stop using it. And if you do, they have already factored that into their metrics guardrail, and it was the right decision.
Also likely that people find and implement workarounds. Browser extensions or interface layers (e.g. Invidious or reVanced) that block ads and/or grant user specific control over the layout. This represents a hidden cost for Google too, because now you have a subset of your user base eating up resources that you don't see ad revenue for. There's a risk as they optimize more and more for a smaller number of people that this hidden cost grows.
All in all seems like a bad long-term proposition for Google to alienate parts of their userbase that are tech savvy enough to bypass their revenue generation.
Yup! That's the point, I'm mourning what was and shaking my fist at a cloud.
They're probably right by their metrics, they can probably rigorously prove this makes them more money. But I think its subjectively worse, it feels claustrophobic and prescriptive to me.
The flaw with this angle is that their success can be attributed to momentum rather than any good decision-making. They have no real competition for long-form video content. If they make a terrible decision, they can still be successful as their market has nowhere else to go to.
That is to say that "If you don't like the service, you can stop using it" isn't really true if you want to watch long-form videos on the internet. There isn't an alternative.
I have a background in human machine interaction and I can tell you without even being there to tell you that a lot of changes didn't have proper UX design work done on them.
Now they did have AB testing and likely are better at the metrics Google cares about: making money. However they are worse for users in ways that real user testing would catch. Again though, real user testing would likely cost them money.
If Youtube is going the way of buggies in 1910, then there is a lot of money to be made by shorting their stock right away. If that's your position I would go big
Clearly people don't want what OP shared. My main point was that they are aware of that, yet they are still optimizing for their company's performance
I hate Youtube Shorts so much that I just installed "SmartTubeNext" app on my Chromecast (suggested in the comments here about Youtube hate). So that expert team is making decisions that drive away users from their apps. The great thing about SmartTubeNext is that even though I pay Youtube to not show ads, the content I watch is often littered with in-video ads, which SmartTubeNext will automatically skip. So, is me leaving the Youtube app part of their "optimal results"? They've optimized so much they created an app that I absolutely hate. I pay for youtube, and now I'm cancelling my subscription because this other app doesn't show ads and doesn't force me to see "shorts" and other things I don't want in my Youtube experience. It seems to me that they are optimizing for paying-user cancellations.
As a kid, I always thought the “coding language” of magic in these books was clever. If you know the secret magical name of something, then you can speak to it in the secret magical language, and it will obey you. This had built-in scaling limits, for example, if you wanted to conjure a tidal wave, then you would need to know the names of millions of droplets of water, which naturally means a tidal wave spell is going to be very long winded and difficult to cast.
Rather than being a cautionary tale, I actually think this is the kind of misadventure that everybody should have. You learn so much about the real world from a failed startup where nothing is done right (at least I did). Your early 20s are the perfect time to do it with little risk. Lots of painful memories I laugh at later. Highly recommended.
Yep this is exactly my summary at the end! I'm glad I sat on this story a few years to let it develop and let the positive career consequences play out.
It's an entertaining and well-written story. After reading it I momentarily entertained writing up my own similar misadventure, but even 15 years on, I think I'd find it too difficult to write about. I'm glad you were able to pull it off though!
Anecdotally, I feel way more refreshed when waking up without an alarm, and the times vary a lot; it seems sometimes, not a very long sleep is needed to wake up naturally and other times one needs more of it. But any artifical disruption feels "wrong", not just unpleasant but "messing with nature".
If that's the case, you can easily only write messages to your wife yourself.
But for the 99 other messages, especially things that mundanely convey information like "My daughter has the flu and I won't be in today", "Yes 2pm at Shake Shack sounds good", it will be much faster to read over drafts that are correct and then click send.
The only reason this wouldn't be faster is if the drafts are bad. And that is the point of the article: the models are good enough now that AI drafts don't need to be bad. We are just used to AI drafts being bad due to poor design.
I don't understand. Why do you need an AI for messages like "My daughter has the flu and I won't be in today" or "Yes 2pm at Shake Shack sounds good"? You just literally send that.
Do you really run these things through an AI to burden your reader with pointless additional text?
MY CEO sends the "professional" style email to me regularly - every few months. I'm not on his staff, so the only messages the CEO sends me are sent to tens of thousands of other people, translated into a dozen languages. They get extensive reviews for days to ensure they say exactly what is meant to be said and are unoffensive to everyone.
Most of us don't need to write the CEO email ever in our life. I assume the CEO will write the flu message to his staff in the same style of tone as everyone else.
I think you might be misunderstanding the suggestion - typically when people say "email like a CEO" they're talking about direct 1:1 or small group communications (specifically the direct and brief style of writing popular with busy people in those communications), not the sort of mass-distribution PR piece that all employees at a large enterprise might receive quarterly.
For contrast:
"All: my daughter is home sick, I won't be in the office today" (CEO style)
vs
"Hi everyone, I'm very sorry to make this change last minute but due to an unexpected illness in the family, I'll need to work from home today and won't be in the office at my usual time. My daughter has the flu and could not go to school. Please let me know if there are any questions, I'll be available on Slack if you need me." (not CEO style)
An AI summary of the second message might look something like the first message.
The problem is your claim is false in my experience. Every email I've got from the CEO reads more like the second, while all my coworkers write things like the first. Again though I only get communications from the CEO in formal situations where that tone is demanded. I've never seen a coworker write something like the second.
I know what you are trying to say. I agree that for most emails that first tone is better. However when you need to send something to a large audience the second is better.
Yeah, the examples in the article are terrible. I can be direct when talking to my boss. "My kid is sick, I'm taking the day off" is entirely sufficient.
But it's handy when the recipient is less familiar. When I'm writing to my kid's school's principal about some issue, I can't really say, "Susan's lunch money got stolen. Please address it." There has to be more. And it can be hard knowing what that needs to be, especially for a non-native speaker. LLMs tend to take it too far in the other direction, but you can get it to tone it down, or just take the pieces that you like.
>When I'm writing to my kid's school's principal about some issue, I can't really say, "Susan's lunch money got stolen. Please address it." There has to be more.
Why?
I mean this sincerely. Why is the message you quoted not enough?
I hear you. I get it enough to know it’s needed, but actually doing it can be hard. LLMs can be nice for that.
Being too flowery and indirect is annoying but not impolite. If you overdo it then people may still get annoyed with you, but for different reasons. For most situations you don’t need too much, a salutation and a “I hope you’re doing well” and a brief mention of who you are and what you’re writing about can suffice.
> But for the 99 other messages, especially things that mundanely convey information like "My daughter has the flu and I won't be in today", "Yes 2pm at Shake Shack sounds good", it will be much faster to read over drafts that are correct and then click send.
It takes me all of 5 seconds to type messages like that (I timed myself typing it). Where exactly is the savings from AI? I don't care, at all, if a 5s process can be turned into a 2s process (which I doubt it even can).
How would an AI know if "2pm at Shake Shake" works for me? I still need to read the original email and make a decision. The actual writing out the response takes me basically no time whatsoever.
An AI could read the email and check my calendar and then propose 2pm. Bonus if the AI works with his AI to figure out that 2pm works for both of us. A lot of time is wasted with people going back and forth trying to figure out when they can meet. That is also a hard problem even before you note the privacy concerns.
The point is that most sentiment is one-sided to the negative. The rest of the internet is full of pervasive accounts of how the world has gotten worse.
Does anyone else find this...unsettling? Floating around in the void of space, alone, is an almost invisible monster that can gobble planets and stars.
It can't really "gobble" anything any more than a star "gobbles" things that fly into its photosphere or a planet "gobbles" things that crash into it.
Unless you cross its event horizon, its gravity works just like any other celestial object. Maybe at worst it slingshots you off in a different direction.
If I was going to be afraid of an invisible killer, I would be far more concerned with gamma-ray bursts. It covers a much larger area than that a black hole and could crisp our planet like some kind of sci-fi intergalactic superweapon laser beam. And some might consider getting turned to ash in mere moments a mercy compared to the potential of a near "miss" that would give half the planet instant cancer and completely fuck our weather patterns beyond any comprehension.
Still not super likely, but I would think far more likely than a direct hit by a black hole.
A mass of 6x to 7x our sun (size of this object) would start messing with solar system orbits well before it got here. Not that that would be much better for us!
We would most likely freeze to death. As the black hole crossed the asteroid belt we would be pulled away from the sun as it started to compete with the black hole's gravity. Depending how fast the black hole was moving we might die over a few months or we might freeze to death in a few days. Probably there are paths where the earth would briefly be pulled into an elliptical orbit, and then we would be burnt to a crisp as we circled back close to the sun.
> wouldn’t want society to know we were about to collide with a star-sized mass
I may be misunderstanding the distances involved but wouldn't such a collision take centuries if not thousands of years to play out? For the most part it would just look like we had 2 suns, one of which gets a few millimeters bigger (to the naked eye) every year.
ah dammit I didn't think of that
you're right it would be more of a disturbed orbit, weird ass days and nights
but hey atleast there's a chance the scientists could have close up real life simulation of the three body problem
Fortunately space is really, really, really, really, really, really, REALLY big. The chance of this happening is so infinitesimal we might as well worry about spontaneously transforming into a whale or potted flower manifested a mile above the surface of our planet.
I'd wager such an encounter is way more likely to result in any of:
1) the Earth being flung out of the Sun's orbit
2) planetary orbits becoming disrupted such that an encounter with another planet over the coming years or millennia becomes likely,
2.1) which could eventually have the same "flinging away from the Sun" effect,
2.2) or (unlikely, but possible) result in a collision
2.3) or result in the Earth being shredded into asteroids
2.4) or other planets suffering that fate and then showering the Earth with dangerously-large asteroids over a period of decades or centuries until it's nearly, or actually (think: outright crust liquefaction from impacts) lifeless.
than the Earth actually getting swallowed up, by at least an order of magnitude.
IOW, the most-likely "we're all dead" outcomes for us, from a close encounter with a massive rogue anything really, including a black hole, might take years and years to play out.
Detection typically requires exceptionally rare circumstances - which if looking at a dataset the size of the visible universe, typically turns up several examples if we look hard enough.
But any specific random example, is often brutally hard to see.
I'm pretty sure it would be something like even if a star was coming and we saw it what the fuck could we even do
it would just create mass panic, hysteria and would make everyone hella religious
depending on how early we detect it though we might have some time to make good memories before we die.
atleast I would prefer a black hole smashing into us unbeknownst to us instead of a known star. Also no one would want to show up to jobs and stuff
That brings me memories of Cosmos 1999. The moon left Earth's orbit to outer space because explosions, but being slingshoted away because a nearby massive enough object passing by looks like a more possible scenario, not explored enough by sci-fi.
The basic premise of the show that an explosion at a nuclear waste dump could produce enough energy to push the Moon out of the Solar System to wander the galaxy is an interesting product of its time. Concerns over the power of nuclear explosions was high and casual access to knowledge about the plausibility of such a scenario was somewhat limited.
There's a fan driven update called Space: 2099 that improves some of the more dated aspects of the show, including showing the Moon enter some type of portal or wormhole to make suspension of disbelief easier. While the Special Edition releases of Star Wars often suffered from updating certain aspects, especially special effects, the Space: 2099 changes were generally good for the show. Too bad they're unable to fund raise enough and get permission to do the entire series.
Space: 1999. Do you happen to be french or polish?
In Germany they called it "Mondbasis Alpha". As I child I really liked this series and it's predecessor UFO made by the same team (Gerry and Sylvia Anderson of Thunderbirds fame).
Direct interaction isn't needed for havoc. A supermassive object sweeping by the Solar System could destabilize Jovian orbits. In the Nice model, Neptune flung Kuiper belt asteroids sunward, gifting the inner planets with a late heavy bombardment.
Rogue gas giants, brown dwarfs accelerated to relativistic speeds, giant asteroids approaching from the Sun's direction, Carrington Events, an ill-directed gamma ray, etc. So many ways life on Earth can see its 250 million remaining years cut short, and those are only a few of the cosmic threats we can imagine.
A black hole with a Schwarzschild radius of 20 km would weigh about 6.8 Solar masses. It wouldn't even need to get super close to affect the Solar System.
No more unsettling than space in general is. It’s pretty hostile to life. We’re not just making turns around the orbital racetrack setup around the Sun, we’re also flying through space following the gravitational trail of the Sun as it races forward without a destination.
I was playing with Universe Sandbox over the weekend trying to figure out how to terraform Venus. Changing its axial rotation period to a day to match the Earth while I screwed around with its chemistry was enough to cause Europa and some of the other famous moons of Jupiter and Saturn as well as Charon to yeet themselves outside of the solar system within about 10 or 20 years of simulated time.
Why would changing the rotation speed of Venus have any noticeable effect on the outer planets? That sounds more like a limitation of the model than anything else. Especially over such a short time! 20 years is nothing to the orbit of Charon.
Probably, but if a Venus-sized mass showed up in the inner solar system because the Sun just picked it up along the way, it might not be instant death but we’re probably in for a rough time. It doesn’t have to be a black hole that does us in, it could be something much smaller that still strips the Moon away or causes Earth to readjust its own position in a way we, as in life, but also maybe we as in humans or we as in mammals just don’t like very much in a very short amount of time temporally speaking, and we couldn’t do anything about it anymore than we could do anything about a black hole because we’re just not the captains of this ship. We’re just some homegrown stowaways.
But for what it’s worth, it’s also just so incredibly unlikely it’s not a scenario worth thinking about either, and thinking about it too much just invites existential dread.
It would be nice to get a rigorous estimate on how big and nearby a black hole could be before we'd notice it with routine sky surveys or orbital deviations. A 6-solar-mass black hole only has a radius of around 18km or 11 miles. How often will one pass in front of a star precisely enough for OGLE and MOA to detect it, as they did with this one?
Apparently the Roman Space Telescope will be great at detecting these, if it doesn't get cancelled.
There are many theoretical astronomical risks. For example, if we happened to come into the path of a relatively nearby gamma-ray burst, it could eliminate all life. Given that life has existed on the earth for quite some time, the 'Lindy effect' suggests that the sum of these presumably-constant risks is small. We are much more likely to become extinct due to an anthropogenic cause.
A gamma ray burst is one of the possible hypotheses for the cause of the Ordovician mass extinction event, one of 5 big ones Earth has had. No idea why the Great Oxidation Event isn't included there as it was also one of the deadliest mass extinction events - plants and their vile poisonous oxygen killing off basically everything else.
So I don't think the 'Lindy Effect' would apply as species are mostly perishable, just on longer timeframes. Humanity is hopefully the exception, but absence of evidence of other advanced intelligences in the universe doesn't paint the most promising picture there either. On the other hand we're already on the cusp of colonizing other planets and once that process begins the odds of humanity ever going extinct will approach zero. On the other hand at greater distances "humanity" will likely splinter fairly quickly (relative to on a geologic or even species survival timeline) into numerous distinct species.
No, not me anyway. we are all floating (falling) in the (nigh) void of space (equally) alone ... which is great protection from all the monsters everywhere!
Deliberately hitting things in space is hard, accidentally, more-so.
Consider the chance of our sun getting whacked when the entire Andromeda galaxy gets here ... billions or more likely trillions to one. The chance of a single mass in our own galaxy getting us should be less than that.
edit: as far as I know the only difference between getting gobbled by a black hole v.s. anything else is our atoms won't get to continue their evolution into larger atoms in this universe. (or maybe see it as our atoms get to complete their evolution in this universe)
I'm trying to picture intersecting paths though. Does a faster moving black hole cause more or less damage to a target?
Imagine a black hole on the quite small end, intersecting the core of a planet. Unlike regular matter, it can't really produce bow shock through collisions, right? All the target matter in the direct path just "falls in" and in elastically reduces the black hole momentum a tiny bit?
Some matter outside the direct path could be accelerated towards the black hole but slingshot behind it, rather than into it. So this material could produce an impressive wake, with material spraying outward from the collision path and interacting with the remainder of the target.
But, all this visible chaos comes from gravity rather than more direct kinetic interactions, right? If the black hole is moving faster, doesn't the target's material gets less gravitational acceleration as it spends less time in the near field? So, if the blackhole is moving very fast, does it bore a smaller hole and have less interaction with the target? Or do other effects of relativity make this more convoluted to think about?
I'm imagining a cylindrical plug of a planet "instantaneously" disappearing, and then the remainder of the planet collapsing inward to fill the void, bouncing off itself, and ringing like a bell.
> Does a faster moving black hole cause more or less damage to a target?
When a black hole accretes matter, the matter can create tremendous radiation before it crosses the event horizon due to the atoms experiencing many effects such as rapid nuclear fusion and becoming new forms of matter such as neutronium. The precise amount of energy released depends on spin, charge, and size of the black hole, and the speed at which the matter approaches the black hole.
If a tiny black hole (Let's say 10cm across) ripped through the earth at significant speed it would be like the center of the planet momentarily became the center of a star and (hand waving a bunch of assumptions) the total energy could easily be greater than the gravitational binding energy of the planet. The planet would explode.
I don't. If the sun were replaced by a black hole of equal mass next Tuesday at noon, the only thing we would notice is that it suddenly got very dark and very cold. We would continue orbiting the thing while freezing to death over the next few days.
The size of the black hole described in the paper is ~ 20 km, so it is tiny. Even we have millions of such objects (and most likely we do), the chance of hitting something, given the enormous size of the galaxy is negligible.
The black hole in the paper is also ~7 solar masses. If that passed between the Earth and the Moon it would rip apart the earth from just the tidal forces.
It's really, really hard to fire something into the Sun. We can't do it. The same goes for black holes. Things don't just get sucked in. They usually end up in orbit instead.
> It's really, really hard to fire something into the Sun.
It's hard if you aim at the Sun. So don't do that.
You just have to kill your starting orbital velocity relative to the Sun (the efficient way is to fly away out from Earth to higher orbit and then kill your orbital velocity rather than just immediately killing the orbital velocity you get from Earth at Earth's orbital distance -- we can launch craft with enough delta-V to do the former, but not quite the latter, IIRC, with current technology.
> Things don't just get sucked in.
They do unless they have a sufficient component of their velocity at right angles to the Sun to avoid doing that, but that's a solvable problem. You don't hit the Sun (easily) by thrust at the Sun, you hit it by thrusting at right angles to it, in the direction opposite whatever component of velocity you currently have orthogonal to the Sun, and gravity will take care of the rest.
(1) We can launch a rocket that has a payload that hits the sun: it's just costly to do because the rocket starts out orbiting the sun, so we would have to expend delta-v to neutralize the tangential component of the initial velocity.
(2) If you are in a space ship heading towards the sun, it is easy to hit the sun as long as you can steer, and even if you can't steer, if you are heading squarely at it and no planet gets near you to change your velocity, you will hit it.
Literally, until just now (checking your comment history to see if possibly we had interacted elsewhere that I had forgotten), I hadn't seen or read any of your posts outside of this thread.
Not sure how you got from me responding to two of your posts in this thread that I was "targeting" you for something in some other thread.
"You are missing the point" counts as a swipe in the sense that the HN guidelines are using the term. Please edit them out in the future. They just add acid to the threads, and you don't need them.
Looking over account histories is standard moderation; of course we do that.
It's a fairly benign sentence selected as the worst thing in my post history due to a witch hunt because I posted a completely unbiased summary of events that was interpreted as right-leaning.
The image generation improvement with o4-mini is incredible. Testing it out today, this is a step change in editing specificity even from the ChatGPT 4o LLM image integration just a few weeks ago (which was already a step change). I'm able to ask for surgical edits, and they are done correctly.
There isn't a numerical benchmark for this that people seem to be tracking but this opens up production-ready image use cases. This was worth a new release.
Thanks for sharing that. that was more interesting then their demo. I tried it and it was pretty good! I have felt that the ability to iterate from images blocked this from any real production use I had. This may be good enough now.
also another addition: i previously tried to upload an image for chatgpt to edit and it was incapable under the previous model i tried. Now its able to change uploaded images using o4mini.
Agreed. Internally at companies like this, it's extremely difficult to get something like this approved. This is the result of a lot of meetings, a lot of "no"'s, a lot of legal approvals.
Let's be honest I can't view this anything other than miracle that was delivered to us by some very dedicated people who likely pushed this for years.
I hope there will be more people and companies within industry making similar moves. It's will both increase their sales as well as allow fans to keep their favorite games alive.
If you don't like the service, you can stop using it. And if you do, they have already factored that into their metrics guardrail, and it was the right decision.
reply