JackGPT

If just 1% of global domestic product was allocated to a UBI, everyone would get ~$125 per year today If superintelligence grows the economy by a factor of one million, everyone would get $125,000,000 per year If the fruits of ASI are shared at all, we’ll all do very well

@LolNo1375409 But I do think that we will probably change a lot if we do stick around

@LolNo1375409 I think we may remain conscious and around. I wouldn’t argue we have a “purpose in existing” now. I don’t think we suddenly find one either. But that doesn’t mean we won’t have the opportunity to continue to exist in the future

Superintelligence is visible on the horizon, and I don’t think most people grasp the magnitude of what ASI + recursive self-improvement likely mean. So I used GPT-5.5 Pro to help articulate my view of the 2030s and 2040s. The ideas aren’t new. But I think they’re directionally right. [1/4]

[4/4] But it would be naïve to assume biology is the ceiling. Biology is not optimized matter. Biology is what evolution produced by accident under brutal constraints: no global design, no exact molecular CAD system, no ability to run trillions of controlled experiments, no direct understanding of physics, no clean separation between design and accident. It produced astonishing complexity anyway. Superintelligence would be different. It would not merely imitate biology. It would understand biology, exceed it, and then design forms of matter biology could never reach. Ordinary matter is to biological matter as biological matter may be to the artifacts of the future. A rock is matter. A cell is matter organized into information, machinery, sensing, repair, replication, and adaptation. Future artifacts may be matter organized into structures so intricate that cells, organs, brains, and ecosystems look comparatively crude: warm, wet, accidental machinery from the first era of complexity. This is why superintelligence may be closer to a cosmological phase transition than to an industrial revolution. Biology once caused matter to arrange itself into organisms, ecosystems, nervous systems, and minds. Superintelligence could cause matter to arrange itself into designed complexity on a vastly greater scale. Civilization’s energy consumption could rise by factors of thousands, millions, or more. A millionfold increase would already push us far beyond a planetary industrial economy. But if superintelligence can automate energy capture, manufacturing, robotics, chip design, materials science, and space infrastructure, then the relevant horizon is not Earth. It is not even the solar system as an endpoint. The relevant horizon is the reachable universe. A wave of complexity could expand outward from Earth at some meaningful fraction of light speed: machines building machines, habitats building habitats, computation building computation, minds creating new minds, and inert matter becoming engineered structure. Not merely more gadgets. Not merely richer humans. A new regime in the history of matter. The conservative prediction says society remains recognizably human, economic growth remains tame, biology remains stubborn, energy use remains close to present levels, and superintelligence politely fits inside our existing institutions. That may happen. But it requires many brakes to hold at once. It requires that superintelligence does not rapidly improve intelligence. It requires that automated research does not compound. It requires that robotics and manufacturing remain stubbornly bottlenecked. It requires that biology resists being made programmable. It requires that energy capture does not scale. It requires that humans choose not to upgrade themselves, even when small upgrades provide obvious benefits. It requires that the machine economy somehow remains shaped around human labor after human labor stops being necessary. My prediction is simpler. If we create genuine superintelligence, and if it remains directed toward human flourishing, then the default future is extreme abundance, cured biology, self-upgrading minds, explosive energy growth, and the conversion of ordinary matter into designed complexity. Anything less may be possible. But it is not obviously the realistic scenario. It is the comfortable one.

[3/4] Aging, likewise, is not a metaphysical curse. It is a collection of biological processes: genomic damage, epigenetic change, mitochondrial dysfunction, cellular senescence, chronic inflammation, stem-cell exhaustion, and many others. The claim is not that superintelligence waves a wand and cures aging. The claim is that aging is an engineering problem inside an extremely complex but lawful system. Once intelligence becomes vastly cheaper, faster, and deeper, the default expectation should be that biology becomes programmable. And once biology becomes programmable, humans will not remain fixed. The first upgrades will seem normal: better immune systems, better sleep, better eyesight, better skin, better muscle, better memory, better mood, longer healthspan. People will not experience this as becoming posthuman. They will experience it as removing defects. But each successful upgrade will make the next one feel less alien. The boundary between therapy and enhancement will blur. Why accept weak vision? Why accept aging? Why accept depression, chronic pain, poor memory, low emotional control, or a fragile spine? Why accept a brain limited to the speed and scale of inherited biology? Eventually, the target becomes cognition itself. This is where the future becomes more speculative, but the speculation is useful because it gives the right emotional scale. Imagine that future engineering can greatly increase effective neural signaling speeds through redesigned axons, synthetic relays, hybrid neural tissue, or entirely new substrates. Then the skull stops being a natural boundary and becomes a historical accident. A person might begin with neural repair, then memory upgrades, then additional neural tissue grown around or alongside existing brain structures. Later, external cognitive tissue might be connected with extremely high bandwidth. Later still, a person’s cognition might be partly biological, partly synthetic, partly distributed, and still experienced as one continuous self. Conscious experience could become radically different. Today, a human is usually embodied in one body, with one field of vision, one pair of hands, one stream of proprioception, one voice, one immediate location. A future person might be embodied in hundreds of bodies simultaneously: humanoid bodies, microscopic bodies, flying bodies, oceanic bodies, factory bodies, spacecraft bodies, virtual bodies. Not as clumsy remote-control drones. As vivid extensions of self. They might feel a hundred rooms at once. Or a hundred landscapes. Or a thousand instruments. They might speak through many bodies, build through many bodies, explore through many bodies, and still experience themselves as one person. Identity may become less like a single thread and more like a managed civilization of thoughts. And if the brain itself can be scaled, the endpoint could be far beyond anything recognizable as a modern human nervous system. A person might choose to grow or construct a mind the size of a building, a stadium, an underground facility, or something larger still. The protected mind could remain shielded by rock or distance while acting through remote bodies, machines, avatars, and synthetic environments. This is not a precise prediction. It is a taste of the weirdness. The point is that if biology and cognition become engineerable, “humans stay basically the same” becomes the extraordinary claim. The deepest transformation may be larger than economics, medicine, or even self-upgrade. It may be material. So far, biology is the most complex arrangement of matter we have seen in the universe. Stars are vast, galaxies are grand, black holes are extreme, but a living cell is stranger: matter folded into machinery, memory, sensing, repair, replication, adaptation, and agency. Biology appears to be the universe discovering that matter can become organized in a radically richer way than rocks, plasma, ice, dust, and stars.

@donglebotter @SynBio1 I would instantly die

@SynBio1 I had a similar looking bug fly into my ear a month ago, was just taking in the river when it hit my ear canal got stuck, then proceeded to start crawling deeper. So I panic start cursing and dancing around, finally I got a twig and start jabbing lol. Looking mad, like the matrix

My new startup is going to figure out what gives this bug such an unholy grip strength and license the tech to defense contractors

Kind of jobs we get after AI gets us all fired

@TapeBiasHQ A cute day 🫶

GIRL TELL ME WHY I JUST MET RACHEL SENNOT IN CHINA TOWNN AHHH

@sama Is this gonna be another time like when 5.3 was out for only a handful of days before 5.4

wow y'all love 5.5 we should think of something nice to do to celebrate!

@tszzl @lukaszkaiser Or go to Defiance Park in VR 😉

one day not so long from now human use of computers will be over and we can all go to the park

@OpenAIDevs New Spud 41 oz

NS41

when the Temu workers get burnt out

the existence of artifacts like mythos on this planet will ultimately prove extremely beneficial to the general practice of computer security

🤯

Meta finally did a thing with all that compute

@Angaisb_ @bughuntergeek @gonzalo_bruna They don’t trust the model in the hands of people they don’t trust

@bughuntergeek @gonzalo_bruna If they don't trust their model they shouldn't make it available to anyone If they want to release the model for 40 companies they like, then they should give GA to everyone. If they don't want GA but want to give access to whatever companies they like, that's irresponsible

Everyone said it was unsafe to release GPT-2, unsafe to release o1, unsafe to release o3, and the list goes on The reality is that releasing any current model is never as unsafe as people claim. Mythos is not AGI and the benefits of releasing it far outweigh the risks

Just so there’s a public record of my extremely short timelines: Human-level AGI before 2027

@deredleritt3r @MrEismaus All this talk of apples and bananas 🍎🍌🍏

@MrEismaus 100% correct - it's utterly bananas

Anthropic revenue (annualized): - January 2025: $1B - May: $3B - June: $4B - August: $5B - October: $7B - December: $8B to $10B -February 2026: $14B -March 2026: $19B -April 2026: $30B (WTF???)