ACM Queue

08/25/2026

Arithmetic Without Numbers

What happens inside an LLM when it tries to calculate with nothing but matrices

How does a large language model do arithmetic when all it has are matrices—no fingers, no scratch paper, no columns of digits? This article looks inside a frozen Llama and asks a sharper question than whether the model can call a calculator: with the prompt text removed, can its own activations reveal the operation and operands? In these experiments, they can. Under a strict no-parser rule, activation-derived readouts supplied a calculator’s arguments for several arithmetic tasks. The same activations reveal number-related directions and rotations, a geometry closer to a clock or spiral than to written columns. Interventions test whether selected internal states affect behavior, while provenance audits trace the calculator arguments back to model activations. Together, the results show one way arithmetic structure can take shape inside a language model.

08/18/2026

Too Many DevEx Metrics, Too Little Guidance

DevEx Metrics Compass: making sense of developer experience measurement

As AI-augmented development becomes standard practice, engineering leaders face mounting pressure to demonstrate its impact. Yet most organizations are measuring AI adoption and output while the effect on developer experience (DevEx) remains largely unknown. The right metrics can close that gap, surfacing the everyday friction developers face and monitoring the real impact of initiatives such as AI rollouts. But with more than 120 metrics in circulation across frameworks, companies, and scientific literature, identifying the right, meaningful, actionable ones for a specific context has become a challenge in itself.

This article introduces DevEx Metrics Compass, a public, open-source web app built on a structured analysis of that landscape across more than 50 engineering organizations. Alongside the tool, we share what the dataset reveals about how DevEx is measured today and where the gaps lie. Whether you are new to DevEx measurement and looking for a first meaningful set of metrics, or a seasoned practitioner assessing whether an existing collection has the right breadth and depth, Compass helps teams navigate the fragmented landscape and identify measures that actually matter for their context and goals.

As AI-augmented development becomes standard practice, engineering leaders face mounting pressure to demonstrate its impact. Yet most organizations are measuring AI adoption and output while the effect on developer experience (DevEx) remains largely unknown. The right metrics can close that gap, sur...

08/14/2026

The Challenge of Efficiency Versus Isolation:

Distributed systems lessons from serving inference

The mechanisms that make production inference affordable are the same ones that dissolve the isolation and recovery guarantees built over two decades. This is a property of the workload, not a failure of the people running it.

The mechanisms that make production inference affordable are the same ones that dissolve the isolation and recovery guarantees built over two decades. This is a property of the workload, not a failure of the people running it.

08/11/2026

Operations and Life
Unethical Ways to Manage Technical Debt

You can ignore it, tie it to unrelated company initiatives, or feign ignorance and simply not offer solutions that add to it. Or… you can adopt more above-board techniques such as Crawl/Walk/Run and the Reverse Farmer’s Rule, which reframe projects so that confronting tech debt is an upfront part of the plan.

You can ignore it, tie it to unrelated company initiatives, or feign ignorance and simply not offer solutions that add to it. Or… you can adopt more above-board techniques such as Crawl/Walk/Run and the Reverse Farmer’s Rule, which reframe projects so that confronting tech debt is an upfront par...

08/06/2026

Where to Draw the Line
- Craig A. N. Soules
Which decisions should not be handed over to LLMs?

In the teams that have leaned hardest into AI, models already write most, and sometimes all, of the code. Increasingly, no human reads that code before it ships. I argue that this inevitable automation is only replacing the visible artifact of software engineering, not the actual job. What was never automatable in the first place—deciding what to build, judging whether the result serves its purpose, and answering when it does not—now becomes the entire role. Drawing on observations from teams furthest along this path, in this article I describe the new discipline that emerges: one built on behavioral specifications, engineered dissent, and continuous instrumented oversight rather than authorship. I also confront an uncomfortable consequence: We are demanding senior judgment while eliminating the work that produces it. Finally, I argue there is a class of judgment we should refuse to delegate, even to a machine that appears capable of it.

In the teams that have leaned hardest into AI, models already write most, and sometimes all, of the code. Increasingly, no human reads that code before it ships. I argue that this inevitable automation is only replacing the visible artifact of software engineering, not the actual job. What was never...

08/03/2026

The Coding Philosopher

Indirected Reality:

Language, truth, and LLMs

How can LLMs appear to know facts about the world when they are trained only on patterns in language rather than direct experience of reality? Hallucinations shouldn’t surprise us. The real mystery is how these systems get anything right at all.

How can LLMs appear to know facts about the world when they are trained only on patterns in language rather than direct experience of reality? Hallucinations shouldn’t surprise us. The real mystery is how these systems get anything right at all.

07/28/2026

Where Does the Foundation Come From?
AI can compress, but not replace.
- Ed Solovey

AI-assisted development is now the default expectation for new software engineers, but the productivity it delivers depends on professional judgment, and AI is taking over exactly the work that used to develop that judgment. Drawing on two vantage points—teaching software engineering at Boston University and leading an internship program at Digits, an AI-forward financial technology (fintech) startup—in this article I trace what happened when new graduates were handed the same agentic tooling as senior engineers and why the fix was a phased on ramp that compressed a decades-long career arc into weeks. For engineers with grounding, AI is a multiplier; without it, AI creates a productivity illusion in which output outpaces understanding. The article argues that universities and companies share one obligation: Embrace AI in education and onboarding, but introduce it in an intentional sequence, and assess the understanding polished output cannot prove, so the foundation gets built rather than bypassed.

AI-assisted development is now the default expectation for new software engineers, but the productivity it delivers depends on professional judgment, and AI is taking over exactly the work that used to develop that judgment. Drawing on two vantage points—teaching software engineering at Boston Uni...

07/23/2026

Titan Transients and LLM Scalability:
The ghost in the machine

While training LLMs, OpenAI observed what it termed the compute-efficient frontier (CEF), which characterizes a lower bound on token-prediction error or test loss. This limit has raised concerns that it might constitute a universal constraint on the scalability of GenAI systems. This article presents a model of LLM computational dynamics, based on the universal scalability law (USL), that provides a framework for understanding the CEF. The USL defines bistable minima in the tokenized neural net landscape. The deeper minimum determines the lowest achievable loss of an LLM instance. Successively larger LLMs possess deeper global minima that align with the power-law slope of the CEF. This suggests the CEF is a consequence of larger LLMs detecting inter-corpora correlations on all size scales.

- Neil J. Gunther

07/21/2026

Beyond Zero:
Enterprise Security for the AI Era

The rise of autonomous AI agents and the accelerating velocity of corporate data access are stretching the application-centric model of zero trust security to its breaking point. This article introduces Beyond Zero, a new security paradigm designed for the AI era. The Beyond Zero architecture performs per-resource access decisions for humans and agents at machine speed. By shrinking the trust boundary from the application level to individual action and by coupling static authorization guarantees with dynamic, AI-driven reasoning, Beyond Zero enables a self-defending enterprise capable of mediating thousands of human and machine decisions per second. Google’s vision for the future of this access model as well as a call for industry collaboration and standards development are outlined here.

The rise of autonomous AI agents and the accelerating velocity of corporate data access are stretching the application-centric model of zero trust security to its breaking point. This article introduces Beyond Zero, a new security paradigm designed for the AI era. The Beyond Zero architecture perfor...

07/14/2026

Goodbye, and thanks for all the Bikesheds!

The only bit left of FOSS as we know it will be that users will be able to read the source code. As reproducible builds catch on, they might even be able to compile that, as long as they make no modifications.

Almost 20 years ago, my friend George (aka Kode Vicious) casually mentioned in an email that acmqueue was looking for articles about flash memory. Not thinking through the consequences, I asked, “How many words?”