threatofrain

The disabled are a concentrated special interest group, and we’d always want to keep it that way. But who will nerd out for the disabled if not for those among them who seek disproportionate impact relative to their base numbers?

That’s not bias if it’s more explanatory of the future.

This very might be the only real solution to the cat and mouse game.

You should be running your own benchmarks often given the nature of these products. Releasing these models whenever with whatever corresponding benchmarks provides you no assurance that things will stay the same.

Having a great time with an ice cream brand at a store is no assurance against hollowing out the brand tomorrow.

Chess is an often-used generic translation word for chess-like games. A native speaker might abbreviate and simply say play chess rather than say the full name, relying on the context for understanding.

I don't think this will mean holding back, as nothing will hold back whatever is providing tech to the myriad of parties interested in poking holes. Instead this points to common defense, as anyone who doesn't have access to defensive tech will be toast.

Yes but they obtained this information based on the one-sided courtesy of another party, under a scheme that tried to get as many people to safeguard together.

If the first party to receive this information says, "Wow, of course I must act on this, because customers are on the line," then it makes sense to simply choose who you tell more carefully or not at all.

> I’ve found that LLMs are HORRIBLE at writing decent abstractions, regardless of how much prompting and skills I’ve given it. It will always get the job done, but it will often be unmaintainable and require bigger and heavier refactors as I introduce more and more requirement changes.

I quote this just to put extra weight on this! It's surprising, the effect size of this statement.

I'm not sure why you think Meta would take such a heavy approach, as opposed to just the right amount at the right time for the minimal price and the least commitment.

Apple responds by waiting and partnering with companies like Google, and treating this as a layer that can be swapped out.

Google, Apple, and Microsoft are each in a stronger position to win consumer AI because they're guardians over the foundations of the user's experience. Meta always has to wonder whether they're going to be edged out, like when Apple decided to limit permissions or set new rules on user data handling.

How has Meta responded? With a failed phone and speculative headset.

Apple enlisted Hermes to try and convince watch people that they understand luxury. The leather wrist strap costs as much as a new Apple Watch. As another noted here, premium vs luxury can be quite a step.

As impressive as Facebook paying for your social media experience.

Cloudflare is making aggressive bets as a company, releasing products at a somewhat insane rate. We have yet to see if they’ll stick their landing.

I also have my own doubts about this acquisition, but as I explore sqlite-per-user models I'm also seeing benefits to having some central control plane. I thought they would've made a natural fit on Cloudflare.

I find this release relieving because it fills a hole that was really needed. Another hole would be a proper database and improvements to D1.

We can think of the split between DB and application in terms of DX or code, or we can think of it in terms of colocation of data and compute. The latter case will be compelling sometimes.

By the time you are 18 and going to college it’s time to reap your lifetime of prep, you’re about to spend a ton of money to specialize. What does it mean to show you all the doors and motivations of life when you’re basically going “all in” on a domain?

Hopefully it doesn’t mean extending high school two years into college, still prepping to make a bet not yet made.

If I found out this metaphor for neighborhood windows was actually a window into massive amounts of institutional data, then no, I don't want someone to talk me out of getting more security on my kitchen window.

Security by obscurity is just a bet on weights, a belief that the economic motivations for attacking are insufficient. That worked before, but developments in ML calls to account all the debt we’ve accumulated through that practice.

I don't think your dichotomy works. When an LLM is reaching into agents.md it is absolutely modifying the specs. Who cares about original providence when it ends up in agents.md?

We were feasting on lies. Rather than demand honesty our instincts instead reach for a deathly sort of deliciousness. Anyways what was your point in the discussion?

And what framework would there be for international cooperation? Without that element, this would just be self-destruction in a global race.

I don't really agree that ML killed their business. I paid for their product early on before it was even done but they never really developed their product. I held out hope to the end of a year and ended up never using my purchase.

Compare that to shadcn, which had visible growing pains and yet reached rocketship fame. Shadcn is evidence that developers were hungry for yet another library, and it's also evidence that users were willing to tolerate rough code on the ride to maturity.

I don't believe Tailwind's component libraries deserved to win.

AI is about automation. Drones have killed people before. If they automatically kill people is that different? Is your worry that mere automation is not novel enough?

Snipers have killed people before. If snipers are now automated, is the worry that there’s not enough creativity to warrant a discussion past saying “same old, same old”?

Before we discussed how important security was, we got insurance, we made libraries and products, we used compliance software, etc. Except how honest were we about all that stuff? How much risk was actually in the air, and what was keeping us accountable on security in either direction of over or under-investment?

Now a reckoning is here. The potential to be attacked might actually translate to being attacked.

We're talking about AI developing weapons I guess because we're very focused on generative tech, but AI is already a part of weapons systems today.

OpenAI is willing to credit so plagiarism is not the right framing here.

That's by design, of course. Whether people got smarter or dumber we might still assign 100 to be the median.

Scratch's problem is ecosystem. It develops UX for children who are still early to the keyboard but children learn so fast, and at some point they get past the high friction stage and end up in a desert.