davidpapermill

I actually think we're in a strange situation with AI compute.

Right now, we have models that are statistical models of language, with a world model and reasoning "falling out" of a lot of effort.

It's like we've made something that's a little bit intelligent, and now we're trying to amplify that trick to create something that's quite intelligent. And - don't get me wrong - it works.

But it's also super, super inefficient. We're having machines "think out loud" to compensate for the quality of their thought processes. We elongate the path to make up for the progress made on a given step.

I tink there's probably a much smarter way of doing things that will require qualitative architectural (and quite possibly hardware) innovations. Right now we're on the path to a Dyson sphere: that's probably not going to be necessary once we figure out a smarter way to think.

I've repeatedly bounced on and off 20VC by Harry Stebbings, but this year I'm finally hooked.

The main draw is the episode released towards the end of the week with Jason Lemkin (SaaStr) and Rory O'Driscoll (Scale Venture Partners). With the pace of AI announcements, it's been a good place to recap and analyse the week's events. In particular, Rory's insights are usually spot-on.

Fantastic. How did you learn Clojure? I'm a bit of a fan.

I don't think so. I think Tesla merger with SpaceX, which has the Cursor team and reportedly working on foundation model there.

I imagine the EU would block any attempted takeover of Mistral given recent Anthropic and US govt actions.

Only people with billions of dollars can train foundation models, yes.

But a competitor to Anthropic at the product level? With open source models, very little barrier.

Are the trains located where the urban areas are, or are the urban areas built around the train network?

It's chicken and egg question, but in Manchester and London it's very clear that mass transit led to urban development, rather than the other way around.

It's very surprising that cities like Leeds have no mass transit at all, and sizeable cities like Liverpool and Birmingham don't have much.

Click and hold the back button.

> A huge part of the job of Software Engineering is producing the right amount of code at the right time.

I'd go further and say that usually the goal is to use as little code as possible without sacrificing readability.

Brevity is compression, and compression surfaces the salient points of a problem.

Elegance often comes down to brevity.

It's a good question. We had a previous language, JDoc, based in JSON. It covered only part of the functionality of the XML-based language and was really only for machine-machine.

We researched a bunch of others: languages like LaTeX and Typst are obvious alternatives. We also considered a super-augmented version of Markdown. Even looked at YAML.

Last year we chose XML as the basis for our document language.

It's been a good choice for designing a new language, but we've been really surprised by the poor quality of the available parsers. We figured it would be a solved problem, but we'll be writing our own at some point.

I think with recent changes they still retain data on Enterprise, no? Or have I misread this?

> or automating abusive or spam comments on social media.

Actually, the biggest problem is the automation of inane comments on X. Which is admittedly quite surprising - I would have agreed with OpenAI at the time.

I'd rather trust Marijn's design skills and extensive experience.

There are good backends available for ProseMirror and other editors. It's not hard to set them up.

"just the UI"

That's a hell of a "just", as anyone working on such projects will tell you. It's super super hard to get this right.

ProseMirror (and presumably Wordgard) just gets so much right.

Marijn: just came here to say that I think ProseMirror is a brilliantly designed project - that you're intent on improving on it is amazing dedication to your craft.

> Rocket Lab has secured commitments for a $3.6 billion bridge loan from Deutsche Bank and Wells Fargo to fund the cash portion of the acquisition.

Given the timing, this seems like a risky move as they'll be issuing debt in mid-2027 to refinance the bridge, at a time the market could be saturated / corrected.

https://www.reuters.com/business/media-telecom/rocket-lab-bu...

$14B (!!!)

https://finance.yahoo.com/news/apple-lazy-ai-strategy-could-...

Planned Capex for 2026:

Amazon $200B

MS $190B

Alphabet $175B-$185B

Meta $115B-135B

The Mag 7 spending around $700B on capex this year, expected around a trillion next year.

That's more than their combined FCF, and they're borrowing to bridge the gap.

I'd just do a quick filter, probably deterministic, then perform a deeper comparison on the selected few.

A better way to reformulate this problem is for the LLM to be tasked with making a _comparative_ judgement between two CVs. This should prove much more reliable, especially if you give it a third “too close to call” option. You can also ask for clear justifications of preference.

Searched HN for something like this.

I've been using it a lot this weekend and it's making endless mistakes, not making sense, losing track of context.

I feel like it changed this week.

Is that real? Imagine they've taken the code out if so, difficult to verify.

Came here to say the same thing. My security researcher friends always point out that security is solved: simply don't build the system and there will be no security threats. But that's not entirely _useful_.

Loved reading the article but it's not a great demonstration of protection against prompt injection. Better would be if the agent were instructed to reply to each email, but never to reveal the secret.

Perhaps round 2?

AI also has the misfortune of coinciding with a very difficult period for young people, including limited job opportunities: the financial crash, COVID, Brexit, the political polarisation in the western world.

In some sense it's probably acting as a lightning rod for resentment, and that resentment is combined with marketing spiel from the model companies alongside well-placed concern over the impact of AI on employment.

> Google is now on their 8th generation TPU

Remarkable that the TPU pre-dates the attention paper. Was a solid bet on energy efficient dense matrix multiplication and has stood the test of time.

There's a major problem with evaluating the result of agentic coding, and "loop engineering" exacerbates it.

Like many crafts (painting, music, etc) the true test of great software is time. Not because of taste, but because software is a living artefact that must evolve.

Only over time does a human developer realise they have "painted themselves into a corner" and chosen the wrong abstractions. And it will be the same with vibe coding (i.e. where code is not architected and reviewed exhaustively by a human).

Because companies won't surface their internal codebases, the most likely signs of such problems with be the reliability of applications and services. I don't know if recent outages are indicative of changes in code practice, or just coincidence, but we'll find out over time.

Agents aren't yet strong general reasoners - they are good at reasoning in the shape of their training corpus - so it's essential that important code is reviewed by human experts.

Should be top comment.

I think there _are_ downsides to using open models.

Quality won't be as good. That's a given, at least with SOTA. But there's more: switching between models means qualitative changes in replies, and in strengths and weaknesses on evals. You're also on the "open source track" and there's no guaranteed upgrade path - the creator may simply stop publishing new models. And the fact that they're mostly from China comes with political, provenance, and governance issues.

I think NVIDIA's models are the most promising, because they can serve as a foundation for new model companies. That's a meta-feature that keeps the proprietary models in check. And NVIDIA's incentives are very likely to remain aligned with open-source: they want to encourage competing solutions that rely on their third party hardware.