tmvphil

I'm optimistic that he will actually be a positive force in reforming how the city operates. I think he is pragmatic in that he understands that efficiency in government administration is something that progressives have insufficiently prioritized. His policies are more populist than I'd prefer, but I think not the crazy socialist fever dream that Rs portray it as. The scariest thing for me is the prospect of active sabotage from the federal level, although I don't know how much they have held back.

I know this in the same way that even though I don't know the exact credence to assign the probability of particular bad effects from global warming, I can confidently say that an increase by a factor of 1000 of the CO2 emissions would be a bad thing. This is not because I have done a simulation, but instead my beliefs are based on the assumption that while concerned experts might be wrong in the details, they are probably not wrong with a gap of 3 orders of magnitude.

Tell me the nuance then. If people have concerns about Kessler syndrome at the starlink scale then why wouldn't something literally 1000x bigger be even more concerning.

See my edit. Just one starcloud would represent an increase in a risk factor of over 300 c.f. status quo. Then multiply that by the number of starclouds you think would be deployed.

Not sure what the slippery slope is here. The linked page imagines a 4km x 4km radiator/solar array. The cross-sectional area of the array is going to be directly proportional to the probability of impacting high velocity space debris. In such an event the amount of debris that would be generated could also scale with the area of the array. This seems bad

How is a multiple square-kilometer radiator not just an inevitable Kessler syndrome disaster?

Edit: Some back of the envelope calculation suggests that the total cross-sectional area of all man-made orbiting satellites is around 55000 m^2. Just one 4km x 4km = 1600000m^2 starcloud would represent an increase by a factor of about 300. That's insane.

I simply do not care if advertisers form an accurate view of my desires and beliefs.

I think for me personally, such a linear breakdown of the design process doesn't work. I might write down "I want to do X, which I think can be accomplished with design Y, which can be broken down into tasks A, B, and C" but after implementing A I realize I actually want X' or need to evolve the design to Y' or that a better next task is actually D which I didn't think of before.

Waterfall might be what you need when dealing with external human clients, but why would you voluntarily impose it on yourself in miniature?

Sorry, I'm going to be critical:

"We follow a strict 5-phase discipline" - So we're doing waterfall again? Does this seem appealing to anyone? The problem is you always get the requirements and spec wrong, and then AI slavishly delivers something that meets spec but doesn't meet the need.

What happens when you get to the end of your process and you are unhappy with the result? Do you throw it out and rewrite the requirements and start from scratch? Do you try to edit the requirements spec and implementation in a coordinated way? Do you throw out the spec and just vibe code? Do you just accept the bad output and try to build a new fix with a new set of requirements on top of it?

(Also the llm authored readme is hard to read for me. Everything is a bullet point or emoji and it is not structured in a way that makes it clear what it is. I didn't even know what a PRD meant until halfway through)

The way it totally disregards the many explicit instructions given in the "four panel" comic strip.

Fundamentally we still have the flat namespace of top level python imports, which is the same as the package name for ~95% of projects, so I'm not sure how they could really change that.

Kind of have one with the missing image benchmark: https://openai.com/index/introducing-gpt-5/#more-honest-resp...

As opposed to a hypothetical scenario where it is legal to participate in illegal transactions?

I don't think a chipotle burrito is actually 1600 calories unless you do something non-standard. Probably 800-1100

I turn that stuff off and just use `q chat` for everything, which actually works very well in my experience.

You can define AGI however you want I suppose, but I would consider it achieved when AI can achieve at least about median human performance on all cognitive tasks. Obviously computers are useful well before this point, but it is clearly meaningful line in the sand, useful enough to merit having a dedicated name like "AGI". Constructed tasks like ARC-AGI simply quantify what everyone can already see, which is that current models cannot be used as a drop-in replacement for humans in most cases.

To me, superintelligence means specifically either dominating us in our highest intellectual accomplishments, i.e. math, science, philosophy or literally dominating us via subordinating or eliminating humans. Neither of these things have happened at all.

> There's already a big meaningful gap between the things AIs can do which humans can't, so why do you only count as "meaningful" the things humans can do which AIs can't?

Where did I say there was nothing meaningful about current capabilities? I'm saying that's what is novel about a claim of "AGI" (as opposed to a claim of "computer does something better than humans", which has been an obviously true statement since the ENIAC) is the ability to do at some level everything a normal human intelligence can do.

According to this presentation at least, ARC-AGI-2 shows that there is a big meaningful gap in fluid intelligence between normal non-genius humans and the best models currently, which seems to indicate we are not "already there".

> A theory that demands we accept consciousness emerging from millennia of flickering abacus beads is not a serious basis for moral consideration; it's a philosophical fantasy.

Just saying "this conclusion feels wrong to me, so I reject the premise" is not a serious argument. Consciousness is weird. How do you know it's not so weird as to be present in flickering abacus beads?

ZX calculus is very interesting framework for doing cutting edge research in error correction and gate compilation, but it seems wildly off base as a means of making quantum computing accessible to a broader audience. Anything beyond the simple "teleportation is like pulling a string" picture is extremely difficult abstract manipulations.

(PhD in experimental QC)

Just compared the time to check on a fairly large project:

- mypy (warm cache) 18s

- ty: 0.5s (and found 3500 errors)

They've done it again.

It's seemed to me that all the productivity gains would be burned up by just making our jobs more and more BS, not be reducing hours worked, just like with previous technology. I expect more meetings, not less work.

I'm a bit stuck on this, maybe you can explain why an LLM would have any difficulty writing REST API calls? Seems like it should be no problem.

They have no qubits at all, "logical" or not. yet. They plan to make millions. It is substantially easier to release a plan for millions of qubits than it is to make even one.

They are currently in the process of creating their own build system: https://github.com/astral-sh/uv/issues/3957#issuecomment-265...

I work in the field. While all players are selling a dream right now, this announcement is even more farcical. Majoranas are still trying to get to the point where they have even one qubit that could be said to exist and whose performance can be quantified.

The majorana approach (compared with more mature technologies like superconducting circuits or trapped ions) is a long game, where there are theoretical reasons to be optimistic, but where experimental reality is so far behind. It might work in the long run, but we're not there yet.

As someone working with it day to day, coming from around 18 years of mostly python, I wish I could say my experience has been great. I find myself constantly battling with the JIT and compilation and recompilation and waiting around all the time (sometimes 10 to 15 minutes for some large projects). Widespread macro usage makes stack traces much harder to read. Lack of formal interfaces means a lot of static checking is not practical. Pkg.jl is also not great, version compatibility is kind of tacked on and has odd behavior.

Obviously there are real bright spots too, with speed, multiple dispatch, a relatively flourishing ecosystem, but overall I wouldn't pick it up for something new if given the choice. I'd use Jax or C++ extensions for performance and settle on python for high level, despite its obvious warts.

Since 1800 congress has used its constitutional power to establish agencies. It is congress who has the constitutional power to shut them down, not the president. The president executes the laws passed by Congress.

Only dictionaries. Sets are still unordered. As a person who was just burned.