verdverm

That's essentially what the OpenCode Go plan is half way to, but with clear quotas

There is a difference in that different models consume different a.ounts of resources (electricity), so times are not as uniform as bytes of data

I was almost killed in a crosswalk yesterday by a USPS van, had the distracted driver waited 1-2s more to look up from what ever they were doing, I would be in a real bad way today

The guy I talked to at the office said they have tracking, he could see the incident in near real time, but also wondered why there were no cameras for incidents like this, to which I replied "government tracking," and that there are reasonable sounding arguments for both sides

token vendors on the path to unlimited plans like mobile data, you love to see it

nice, one thing you might also do, in your wasm version, is figure out how to replicate scroll momentum, that's the thing that makes it feel weird, I can flick a page and it will scroll for a while

help people with suicide and school shootings like ChatGPT already has

OpenAi is alledged to have been monitoring these internally and not contacting authorities. Lawsuits have been filed, I see gross negligence without the gory details

I have for more concerns around human-chatbot maladies than I do around the cyber security stuff. For example, why hack grandma when you can get her to do something willingly through impersonation. How do we prove authenticity in a post truth world?

which article? have not seen this one

---

maybe it's this Anthropic post on GLM?

https://www.anthropic.com/research/glm-5-3-and-the-spread-of...

> Governments should conduct safety testing on sufficiently capable AI models, including successors to GLM-5.3. Without high-quality evaluations from independent sources, the impact of these capabilities might not become fully clear to model developers until it is too late. As AI developers across the world build increasingly capable open-weight models, we hope they work to appropriately safeguard these capabilities and prevent misuse.

I for one do not think my government is up to the task of designing or implementing such a system

ah, that explains a lot, I see the attitude resemblence

I'd make a website with regular website technologies, and then create some view/iframe/page that has the usable TUI in the browser (if that's the concept). Keep pages that are mainly text / docs in that standard tech, these oddities and overrides are better when the user knows it's going to happen, or another way

1. landing page, normal tech so you don't turn people off immediately 2. app/tui page, linked to from landing page letting the user know you did some cool rust/wasm/tui thing, don't necessarily have to call out scroll issues, it will be understood/accepted by way of the weird web tech they have clicked through to

I see it on mobile/desktop | chrome/vivaldi

you've done something to the scroll that makes it nearly impossible to move the text and read below

please do not override default scroll behavior

I def wanted to read more, I'm playing around with retro games and agents, using emulators and old terminal games slippedy mind (I also just want to give FF1 a shot first)

cache is typically 10%, is this OAI setting a new level at half, 5%?

I'm not expecting you to human RL here, but I do hope next time you remember there is another imperfect human behind the screen and are less too online in future interactions

https://www.youtube.com/watch?v=c1Fv1uKTd-w

oh-seven

As a developer, a good way to reframe this and step away from your own domain expertise, is to read Terence Tao's conversation with ChatGPT and ask yourself "could I have done that?"

yup, we use Fireworks.ai, and American company with ZDR

Their new Ember-1 model is pretty good, fine-tune of Kimi3 with way less thinking

Does this make it an American model or is it still Chinese? Does it matter?

https://fireworks.ai/blog/ember-1

hear me out, what if we publicly shame them for allowing non-consensual image manipulation like Meta with the perv glasses?

Microsoft "co-perv-it"

it's a shame the way handshake and crypto went, that was an actually great use for blockchain

one of the other feature of PLC compared to did:web is that it presents the identity history, but this is also problematic for some people (dead naming and right to be forgotten)

I believe there is work around did web for an extension that would enable verifiable history

at that point, you can switch to dirt cheap open models

this "feature" is one of the primary that caused me to cancel and move to exclusively open weight based systems

the (Ai) factory must grow!

How are you thinking about technical debt in this new agentic age? (also markdown debt?)

It's one of my bigger concerns and I use this iterative approach to try and reduce it... "go look at the ./cli directory and generate me a report of inconsistencies blah blah..." or "review ./research/something.md, validate consistency, correctness, claims, ..."

... pseudo prose, have we made that a thing yet like pseudo code?

did:web is closer to owning your own identity, this is a central database controlled by an organization

The PLC makes it easy for Bluesky to be a custodian and onboard new users

Is this an attempt to prompt inject a monero transfer? Only darwinian culling candidates would have their keys and agents together

This certainly does not belong on HN

atproto is currently permissionless world readable, that seems worse to most people (I asked low 100s) than being in a siloed platform with some amount of control over who can see their content

oh, maybe you mean Sam Altman's World with their eye scanning for identity, that is quite dystopian

this is human writing...

this is claude writing...

corporate needs you to find the difference

Are you familiar with Mom's for Liberty? https://www.momsforliberty.org/

They have a whole effort around conservative activism through school board membership. They love their book bans.

> exercise violence in a very civilized way

there is nothing civilized about violence

memorizing programming language syntax is a more apt analogy that enabled you to think in higher level concepts when developing software

both extremes have really bad takes, the middle path is the way

The underlying basis for current "conservatives inspire more violence" is largely drawn from clips taken from the manosphere-podcast realm. There is an amount of war hawks always being on the right that predates all of this.

I've seen similar from the left wing, though they don't have good podcast game or modern communication generally.

How much of your preferred programming language have you memorized?

How has this enabled you to engage in higher-level thinking instead of syntax correctness?

(I saw this pattern first hand with 100s of students while teaching intro cs)

---

my current favorite definition for The Singularity is the point at which humans give up critical thinking and have decided deferring to Ai is the better idea, social media being a stepping stone on that path that primed us for the outsourcing

much of the commentary under this submission has me thinking we are well on our way in this journey

ah yea, that is the one!

thanks for surfacing this for everyone

link: https://github.com/mattpocock/skills/blob/main/skills/engine...

let the agents debug these things, without needing images, the feedback loop is key

I haven't done any mobile stuff in a while, but there should be enough around to wire the agents up into a good environment, like remote chrome-devtools (with or without the MCP).

pnpm workspaces, makes having multiple packages with internal dependencies (like core and components) easy to manage in the dev workspace