"Poison" can also refer to a substance toxic to other animals. We say that chocolate is poisonous to dogs for instance. And a good fraction of Earth's biosphere was killed off by oxygen poisoning in the first of Earth's great mass extinctions.
Also, the dose makes the poison and excess oxygen actually can poison humans. Deep sea divers have to worry about excess oxygen inducing seizures if they mess up their breathing gasses enough. And even 100% oxygen at regular pressure will slowly damage the lungs, something ICUs have to worry about.
Nick Lane had a great book about oxygen, Oxygen, which maybe isn't as good as his book about mitochondria but is well worth reading.
"Our descendants are going to enjoy an enormous wealth of imagery and videos for events that will to them otherwise be just something from a history book. "
The question will be at some point, will they be able to tell it apart from AI generated fake ones? (and will they care?)
Already now youtube recommends me some obvious AI generated garbage as WW2 documentations.
And that was just garbage generated for attention (ad money).
Once big actors with money want to rewrite history and flood the web with fake images to spread certain narratives, then new challenges will arise.
I hope enough people still care about facts and guard them.
That's a good point. When I wrote my comment only my optimistic side was engaged ;-). The pessimistic side shares your concerns. I hope that we develop some technologically diffult-to-overcome solutions for preserving the integrity of media. Like methods for cryptographically signing raw content from a digital camera that guarantees it was produced by that hardware. Not a panacea, but a step in the right direction I think.
It's the "if you think the news is all lies, bullshit and agendas you should see the history books" meme.
Lord knows what falsehoods of today will become the official record of tomorrow never mind what lies of the past we just repeat because they're what got written down.
Without massive terraforming all of Mars is very hostile.
But having solid ground is still nice.
A workable compromise is making big habitats in a dome, that gives sunlight, but shields from radiation. And the ground needs to be processed obviously.
Gravity kind of cuts both ways. Closer to that of Earth is nearly guaranteed to be better for long term human health, but there's a possibility that martian gravity is "good enough" when supplemented with excercise while also making heavy operations and getting back out of the planet's gravity well easier.
This is the thinking of someone on the timescales of a single life. If humanity persists another 1000 years on our current trajectory (US/world politics not withstanding), I think nothing is really a fantasy. Rather, it's all possible but maybe just not in our own lifetimes. But it is also terribly difficult for us to plan for tomorrow, let alone for a future where our descendants are at the helm.
I agree, it’s just a failure of imagination. Some folks correctly foresee not being able to continue what we’re doing now in the exact same way in some new context and conclude everything is impossible. Life isn’t this fickle, it’s adapted before and will adapt again. This is why great science fiction is so valuable, as some people are better at imagining new ways of being more than others, and can show the rest of us the possibilities.
The counterargument is a simple opportunity cost calculation:
There will never, ever, ever[1] be a scenario where if you weighed up the options of "expand into some less habitable area of the Earth" versus "expand to Mars", the latter is the better option either 1) financially, or b) quality of life.
Nobody[2] ever picks the dramatically more expensive and dramatically worse option!
Also, people that are desperate enough to even consider living in the least desirable -- but still just barely habitable -- parts of planet Earth are essentially by definition too poor to afford interplanetary travel.
And no, no amount future sci-fi technology can possibly overcome the simple energy costs of this! If someone can afford the hugely energy intensive interplanetary travel, and the up-front investment required to survive incredibly harsh environments, then by definition they could more productively invest that here on Earth! It's the cheaper and better option in every possible way, and always will be.
This will remain true even if it's standing room only on the entire planetary surface -- it'll be cheaper to build levels upwards while digging downwards.
Maybe our atmosphere will become horrifically polluted? Sure, okay, air filters are faaar cheaper than a full vacuum-capable space suit!
Etc, etc, etc...
[1] Okay, fine, maybe in a million years. Whatever ends up preferring Mars at that point will no longer be "human" by any sane classification.
[2] For some values of nobody. There are morons that buy overpriced branded handbags made of literal trash. I doubt idiots like that will make for a successful, self-sustaining colony.
The Madrid Protocol says you can't do anything fun with Antarctica. Can't have a mine, a garbage heap, or a farm. I suppose the world's militaries stand ready to capture any enterprising colonists and destroy their structures.
> If humanity persists another 1000 years on our current trajectory
It's unlikely that we can persist in our current trajectory for another 100 years without catastrophic climate events puttung a stop to all of these endeavours.
Venus seems like a wonderful place to live, relatively speaking.
At the right altitude where you can "float" on the ocean, it's a pretty comfortable temperature and there's plenty of solar energy but you're shielded from the solar radiation. So, long term, your body will still work, assuming you can solve "the other problems."
Of course, the down-side is that there's nothing to stand on and probably more importantly, there aren't many useful materials to work with besides tons of carbon, oxygen, and nitrogen. Not much hydrogen there, so not much water, which probably is the biggest problem. One of them, anyhow. Also, there's probably not a whole lot to do besides float (zoom, actually) around and slowly go stir crazy in your bubble.
But relatively speaking, it's way nicer than living in a hole on mars where you'll slowly die from gravity sickness, or radiation poisoning, or whatever.
> Not much hydrogen there, so not much water, which probably is the biggest problem.
Actually, the cloud layer at that level is mostly sulfuric acid, from which you can get your water. It also means you need to be in a hazmat suit when you walk outside, but that's still a step up from everywhere else, where you need a bulky pressure suit instead.
Whether it is really possible, is a different question, but after you have an atmosphere, you could have engineered microorganism processing the soil etc.
To put it into perspective, we are effectively terraforming Earth today, though maybe not in a good way.
We have converted most of the land to agriculture and released maybe trillions of tons of greenhouse gas in the atmosphere, there are 8 billions of us working on it. And what did we do? Increased the global temperature 2 degrees? Made the sea level rise a couple of meters?
It may be bad for us, but compared to terraforming a planet like Mars, that's nothing, and we have the entire humanity industrial complex to do it while on mars, we need to build everything, starting from a hostile environment.
That's exactly my point. We don't have the manpower, the materials, the industry.
For mars this just isn't happening unless we ship half of Earth's people and resources over there. Who will have to live on a toxic planet.
But we can't even ship all that stuff there because we don't have enough fuel to do that (it requires many times the payload in fuel) and all the launches would make earth uninhabitable. Terraforming mars is therefore science fiction unless we break a lot of barriers like clean fusion, space elevators etc. And even then the material question will remain a problem.
I think even reverting climate change on earth, a much easier problem than terraforming a remote planet, is a pipe dream. If we're going to be going carbon capture at that global scale, we're going to need to extract so much material, manufacture so much equipment, transporter it all, deal with all the captured carbon, maintenance, power etc all stuff that's not possible to do completely carbon neutral, that we're just polluting a lot more. Especially if we want to do it at a timescale where it still matters.
It's just that terraforming will require a lot of materials that will have to be brought over from Earth. And every tonne of materials to Mars requires many tonnes of fuel to launch from Earth.
I don't think it is possible to ever transport enough to make this happen.
I wouldn’t go that far. It was pretty clear a long time ago that humans spending so much time filling the internet with content was going to eventually enable neural networks to pretend to communicate.
The advancements required to arrive at modern LLMs and the tech needed to get humans safely to Mars or live safely on the Moon are orders of magnitude in difference.
If you can kick off self-sustaining biological processes it’ll happen on its own eventually, but you’d just be looking at generational time scales to do it.
Of course you’ll probably have lots of side-effects.
How do we do that? I imagine dumping Earth life on Mars it will just die. What if we buried a terrarium at the Martian pole with a radio isotope and solar heater and controls so that it could try growing bacteria inside and controlled-leaking some outside into a nearby warm (liquid water) surroundings, and that could get many chances to evolve strains that could survive further away - analogous to ocean life around deep hydrothermal vents.
> In that sense then the term "terraforming" is on equal footing with alchemy
NASA has proposed using "synthetic biology to take advantage of and improve upon natural perchlorate reducing bacteria. These terrestrial microbes are not directly suitable for off-world use, but their key genes pcrAB and cld...catalyze the reduction of perchlorates to chloride and oxygen" [1].
2) If you have a source of hydrogen: water. Bonus as you don't have to make the dome hold pressure. A layer of water of the right depth will generate the force needed, the structure only needs to keep itself level. The only pressure holding is outside that, enough to keep the water from boiling. And, well, it's water--if it's hit by a rock that isn't too big you'll just have hole in the top layer, easily fixed. The same general idea would work on the Moon but the water is far from transparent if you pile up enough of it and you need a lot of hydrogen.
Well, I did wrote "gives sunlight" and that is a valid reply to it. But ... I would need sunlight actually. That seems somewhat possible with light tubes, but the much nicer solution, a transparent dome to still see mars clouds at day and the stars at night, is indeed not possible with current materials.
Probably just ignorance, but there actually was a gap in that time in some areas in germany. Close to my hometown are the remains of an old ancient fortress - that was build by mostly unknown people and abandoned at 400 b.c. and only 1000 years later there were settlements again. A bit rougher area, though.
The flat area of Berlin on the other hand, had human settlement since 60 000 years.
I don't think anyone requires every hectare to have been developed in order to consider an area inhabited. Neanderthals are named after the Neander valley (Thal) near Dusseldorf, so I think we can definitely say yes, that humans have lived in Germany since long before antiquity.
"Also notable: 4.7 now defaults to NOT including a human-readable reasoning token summary in the output, you have to add "display": "summarized" to get that"
I did not follow all of this, but wasn't there something about, that those reasoning tokens did not represent internal reasoning, but rather a rough approximation that can be rather misleading, what the model actual does?
The reasoning is the secret sauce. They don't output that. But to let you have some feedback about what is going on, they pass this reasoning through another model that generates a human friendly summary (that actively destroys the signal, which could be copied by competition).
My assumption is the model no longer actually thinks in tokens, but in internal tensors. This is advantageous because it doesn't have to collapse the decision and can simultaneously propogate many concepts per context position.
I would expect to see a significant wall clock improvement if that was the case - Meta's Coconut paper was ~3x faster than tokenspace chain-of-thought because latents contain a lot more information than individual tokens.
Separately, I think Anthropic are probably the least likely of the big 3 to release a model that uses latent-space reasoning, because it's a clear step down in the ability to audit CoT. There has even been some discussion that they accidentally "exposed" the Mythos CoT to RL [0] - I don't see how you would apply a reward function to latent space reasoning tokens.
There’s also a paper [0] from many well known researchers that serves as a kind of informal agreement not to make the CoT unmonitorable via RL or neuralese. I also don’t think Anthropic researchers would break this “contract”.
> If that's true, then we're following the timeline
Literally just a citation of Meta's Coconut paper[1].
Notice the 2027 folk's contribution to the prediction is that this will have been implemented by "thousands of Agent-2 automated researchers...making major algorithmic advances".
So, considering that the discussion of latent space reasoning dates back to 2022[2] through CoT unfaithfulness, looped transformers, using diffusion for refining latent space thoughts, etc, etc, all published before ai 2027, it seems like to be "following the timeline of ai-2027" we'd actually need to verify that not only was this happening, but that it was implemented by major algorithmic advances made by thousands of automated researchers, otherwise they don't seem to have made a contribution here.
> For example, perhaps models will be trained to think in artificial languages that are more efficient than natural language but difficult for humans to interpret.
The first 500 or so tokens are raw thinking output, then the summarizer kicks in for longer thinking traces. Sometimes longer thinking traces leak through, or the summarizer model (i.e. Claude Haiku) refuses to summarize them and includes a direct quote of the passage which it won't summarize. Summarizer prompt can be viewed [here](https://xcancel.com/lilyofashwood/status/2027812323910353105...), among other places.
Are you sure? It would be great to get official/semi-official validation that thinking is or is not resolved to a token embedding value in the context.
Although it's more likely they are protecting secret sauce in this case, I'm wondering if there is an alternate explanation that LLMs reason better when NOT trying to reason with natural language output tokens but rather implement reasoning further upstream in the transformer.
I would doubt it. They are mostly trained on natural language. They may be getting some visual reasoning capability from multi-modal training on video, but their reasoning doesn't seem to generalize much from one domain to another.
Some future AGI, not LLM based, that learns from it's own experience based on sensory feedback (and has non-symbolic feedback paths) presumably would at least learn some non-symbolic reasoning, however effective that may be.
My argument for this is mostly that we don't use language for all forms of reasoning, and are likely doing so on some internal representations or embeddings. Animals also demonstrate abilities to reason with situations without actually having a language.
I see language more as a protocol for inter-agent communication (including human-human communication) but it contains a lot of inefficiencies and historical baggage and is not necessarily the optimal representation of ideas within a brain.
'Hey Claude, these tokens are utter unrelated bollocks, but obviously we still want to charge the user for them regardless. Please construct a plausible explanation as to why we should still be able to do that.'
The buisness opportunity is what they are advertising here, communication with lawyers is protected, continue to go pay real lawyers for every question and don't try yourself with AI, that is unfortunately not protected.
reply