icbms and hypersonics exist, theres no defense against the top end anywhere atm.
if you're in a war where people are shooting your satellites out of the sky you're past the point where defenses matter. theres no grey war in space like russia hiring people to blow up stuff in europe and very few states could take out multiple satellites in space.
beyond that if we ever did get to that point, you could develop anti missile tech in space, you effectively have the high ground sitting at the top of the gravity well.
I'm reasonably sure there are more states that could destroy a satellite than hit infrastructure on US soil. And if you try to use an ICBM to hit the US then you risk getting nuked.
> people are shooting your satellites out of the sky
If space powers have any sense at all they'll develop a fleet of satellites that deploy small limpet craft that attach to the enemies' satellites and drag them toward earth.
Literal shooting sounds like mutually assured satellite destruction.
This. Orbital debris are an area denial attack, and since there's only one shared planetary orbit (in aggregate), debris-creating weapons are more like biological weapons, poison gas, and tactical nuclear weapons. Even if you win, you've impeded yourself.
The internationally hoped for outcome should be that China continues scaling their orbital use, to the point where it becomes mutually advantageous to sign a treaty with the US banning orbital debris-creating weapons (and ideally imposing mutually-supported consequences on non-signatories who breach the agreement).
Then everyone would be left with more sane options of disabling-only (microwave, laser, targeted EMP, etc).
There's allegations right now that the model essentially read the work of a human mathematician using AI to work on the problem and OpenAI is presenting his work as that of their model
Allegations that the model plagiarised itself, while reflecting poorly on humans, don't make the AI any less impressive. It was the one doing the breakthrough on both sides, after all, not the human prompters.
That is how the PR reads, but is not at all what happened.
A team of highly trained and skilled people used an AI tool, through many many instructions (prompts), to produce a specific mathematical theorem. The tool is impressive, the result (possibly/probably) interesting, but the PR skips the vital role of the humans (for the usual PR reasons).
Not to mention that the exponential plummeting cost of tokens means that that $15 million will be a "pocket change" within a decade or less: https://a16z.com/llmflation-llm-inference-cost/
what we're going through is the same thing as smartphones, the limiter is compute.
it used to be snapdragon came out HTC rushed out a janky phone everyone went omg htc is goat, then in the next few weeks and months others would impliment better versions and people would not notice those as much, finally sony would release a polished phone right as the next snapdragon cycle came.
eventually compute gains leveled off and apple won on taste.
nvidia/tpu is the new snapdragon. Anthropic and google both peaked on the first training run on a new tpu cycle.
you should expect amazing things within a few months of each other from everyone with access to chips and willingness to use them on a training run.
We haven't seen willingness from google to do that. So its currently xai,oai,anthropic, and probably soon meta.
if all it could do was operate a lawn mower and a shovel, the annual value to a homeowner in the north east is ~1-2k depending on lawn size assuming they value their free time at the same amount as their hourly wage.
what you have to keep in mind is homeowners salary is way above avg salary so anything you can do to effect them has a much larger impact than it would seem at first.
they're subsidized if you max them out, i'd imagine most users are paying $20 for maybe $2-5 of tokens.
anthropic probably has more customers that use more of their sub, but for open ai where a lot of their subs are consumers through chatgpt.com, they have a lot of free money to work with there
pi is the neovim of agentic harnesses, its barebones and extremely configurable. if you're the sort of person who likes that sort of things its a forever product, nothing is going to displace it because you have full control.
opencode builds a lot more in, which is better if you dont want to fiddle with config.
nice. i had thought the consensus had moved pretty firmly towards pi, so i was surprised to see Thinking Machines demoing their new model Inkling in OpenCode. wondering if they are previewing an acquisition
Agree, after spending too much time and tokens configuring Pi and adding extensions to match other harnesses, I switched to OpenCode and left the Pi customization circle jerk. I have other things to do and IMHO harness engineers should do the harness engineering, I don’t want to waste tokens and time to build and benchmark extensions. Pi is great, but would be better with a set of official, trustworthy and efficient extensions, and opt-in to enable it.
if you're in a war where people are shooting your satellites out of the sky you're past the point where defenses matter. theres no grey war in space like russia hiring people to blow up stuff in europe and very few states could take out multiple satellites in space.
beyond that if we ever did get to that point, you could develop anti missile tech in space, you effectively have the high ground sitting at the top of the gravity well.
reply