标签：#p

Amazon isn't slowing down on data center spending — but investors don't seem to mind.

Apple said a slowdown in mobile gaming and changes to the App Store’s business model — including court-ordered payment rule changes in the U.S. — weighed on its services business, even as the company topped 1.5 billion paid subscriptions.

Apple may allow users to pay to increase their AI usage limits. During an earnings call on Thursday, Apple CEO Tim Cook said that he believes people will want to use Apple Intelligence and the upcoming Siri AI "a lot," adding that "we will have some kind of upgrade possibilities on iCloud Plus where people […]

The iPhone and Mac sold well, but memory costs threaten further price increases.

You are a software engineer. Your craft honed through years of careful practice. Then suddenly, there are these chatbots and agents. Overnight, your colleagues got a new title on LinkedIn: "AI engineer". Some are already SENIOR AI engineers. You're curious about this new world, and might want to catch up and become part of it yourself. If this is you, then join me on this tour through the concepts and patterns that make up the field of AI engineering. We will find that AI application development is mostly 'just' software engineering, applied to one genuinely strange new non-deterministic component: the LLM. During the tour, we build a real application, end to end. Every article adds a new layer. We link the new patterns and words to existing software engineering concepts you already know. Before take-off, I'd like to establish one vocabulary rule used throughout: "the model" means the LLM itself (large language model, like GPT or Claude), and what AI engineers build around it will be referred to as "the application", "the agent" or "the harness". What we're building As I work at a payments company myself, I figured I'd stick to my domain. PayIQ, the application we build, is an assistant for merchants to perform payment operations: issue refunds, defend chargebacks, calculate processing fees. Give it a charge amount and a payment method, and it computes what a refund actually costs (spoiler: more than the refund amount). Ask it whether a chargeback is worth fighting, and it does the expected-value math using your knowledge base. Ask it something it can't responsibly answer, and it asks for what's missing. No guessing, no hallucinations. By the end, PayIQ will have structured outputs that can be consumed by other systems, a tool belt of financial calculators, retrieval over a knowledge base, an agent loop with persistent memory, an orchestration graph with steps the model cannot skip, token streaming behind a FastAPI service, a regression eval suite, and layered injec

Every i18n setup I've seen has the same three-way standoff. Developers want type-safe JSON in the repo. Translators want a familiar tool, not a pull request. Product wants to fix a typo without a deploy. So you either pay $50–$500/month for a localization SaaS, or you copy-paste strings between a translator's spreadsheet and your JSON files until something silently breaks. For projects under ~1,000 keys, there's a better middle: the spreadsheet is the database. Translators edit a Google Sheet; an Apps Script endpoint serves it as clean locale JSON; your app pulls that at build time. Here's the whole pattern, with the code. Why a sheet beats a translation service for small projects A localization SaaS earns its price at scale — dozens of translators, thousands of keys, screenshots and review workflows. A 300-key marketing site doesn't have that problem; it has a coordination problem. A Sheet solves coordination for free: translators already know it, it has revision history and suggested edits built in, and product can change a string in ten seconds. You only add the two things a raw sheet lacks — a clean JSON API and a fallback for missing translations. The schema: one tab, one row per key A strings tab, with the key in column A and one column per locale: key en tr es fr hero.title Welcome Hoş geldiniz Bienvenido Bienvenue hero.cta Get started Başla Empezar Commencer Use dot-notation keys ( hero.title ) so the JSON nests naturally in your i18n library. Keep a tiny meta tab too: B1 = default locale ( en ), B3 = version ( 1.0.0 ). The Apps Script endpoint Deploy this as a Web App (same mechanics as any Apps Script webhook ). doGet serves one locale — or all of them — as JSON, and the fallback lives right in the query: an empty cell resolves to the default locale, so a half-translated key never ships blank. // Code.gs const SHEET_ID = ' your-sheet-id ' ; function doGet ( e ) { const locale = ( e . parameter . locale || ' all ' ). toLowerCase (); const result = buildLoca

TL;DR. Bringing AI coding agents into a German operation is not only a technical decision. The moment they meet real systems and real teams, they also meet three things that international AI content almost never mentions: a data processing agreement, the works council, and the question of who owns the generated code. None of this is a footnote. It decides whether your AI project reaches production or stalls in legal review. Here is the layer most vendors forget, and the fast order to clear it. Why this layer is invisible from outside Germany Most content about AI coding agents comes from a world where a developer tries a tool and starts shipping. In a German company with a works council, GDPR, and a legal department reading along, the path to production looks different. Not harder, but with stops that nobody skips without regretting it later. This is not a brake. It is the actual difference between an impressive demo and something your company is allowed to run. And it is exactly the layer that reveals whether a vendor has ever delivered in Germany or is only demonstrating a tool. The three stops The data processing agreement: who is processing whose data? As soon as personal data is processed on your behalf, Article 28 GDPR requires a data processing agreement with the processor (in German, an Auftragsverarbeitungsvertrag, or AVV). With AI coding agents the decisive question is not "are we using AI" but what does the agent actually touch. Does it run over a codebase that contains customer data? Does it send fragments to a third-party model provider? Who in that chain is the controller, and who is the processor? This is answerable, and it belongs settled before the first access, not after. A clean setup often limits, at the technical level, what data an agent can even see, which makes the data-protection question smaller and the answer simpler. The works council: codetermination over systems that can monitor This is the stop most often missed from outside Germany, b

OpenAI has formally outlined a national science initiative designed to connect frontier AI models with government research infrastructure, National Laboratories, universities, and working scientists. The program is not a single model launch. Instead, it combines funded access, early product access, scientific campaigns, and an emphasis on fitting advanced AI into real research workflows. The initiative gives concrete form to OpenAI’s stated goal of helping scientists use increasingly capable models to accelerate discovery. In its official announcement on advancing the next era of national science , published July 22, 2026, the company describes a long-term strategy built around the U.S. Department of Energy’s Genesis Mission and collaborations with National Laboratories. The core proposition is that AI can contribute to hypothesis testing, simulations, and experimental work when it is deployed alongside scientific infrastructure and human expertise. That framing matters. OpenAI is positioning frontier models as tools that researchers direct and evaluate, rather than as a replacement for the institutions and specialists responsible for scientific work. What OpenAI is providing to scientific researchers OpenAI’s commitments span several types of access, from coding support for a broad research community to model capabilities and API funding for large campaigns. The announced provisions include: $4 million in Codex access for approximately 2,000 Genesis researchers at national labs and universities. $3 million in API support for two large scientific campaigns. Up to $10 million in API usage for participating researchers who reach a $2.5 million spending threshold. Access to GPT-Rosalind’s bioscience capabilities for national-lab researchers. Early access to selected models and features for trusted national-lab leaders preparing workflows and evaluations. Expanded access to advanced cyber capabilities for national-lab cybersecurity researchers. These commitments indicat

TL;DR I designed a proposed AWS Transform Custom community skill that prepares Glue 2.0, 3.0, and 4.0 repositories for Glue 5.0. It separates safe mechanical transformations from changes that require human evidence, generates a migration report, and preserves already-compatible files unchanged. Because I didn't have live atx access, the benchmarks in this post are explicitly labeled manually simulated, not agent-executed. The proposal is open as issue #75 — not yet merged, not yet a pull request. The missing data-engineering transformation AWS Transform Custom can apply agent-driven code transformations across a single repository — or thousands of them at once, via AWS Batch and Fargate. As of July 30, 2026, its public sample repository, aws-samples/aws-transform-custom-samples , contained three community-contributed transformations: an EKS version-upgrade-readiness skill, a JBoss-to-Spring-Boot migration, and a Kubernetes readiness migration. None of them touched data engineering. Given that most of my day-to-day work sits across AWS data engineering, Databricks, and Delta Lake, that gap was the obvious thing to fill. What an AWS Transform Custom "skill" looks like Before writing anything, I studied the deepest existing example, jboss-to-springboot , since the pattern it establishes is effectively an unwritten spec for the other two skills as well: README.md — the problem, what the skill does, and how to invoke it via the atx CLI. This is also where the repo draws a clear line: these are readiness transformations. They modify repository artifacts — code and infrastructure-as-code — but they don't deploy jobs, call AWS APIs to change running resources, or claim data-level equivalence. That distinction matters throughout everything below. SKILL.md — the agent-facing definition: YAML frontmatter with trigger keywords, an Objective, explicit Non-Goals , Constraints, worked before/after examples, a "signal in source code → reference file" routing table, and a numbered V

I just made Whizz, an esoteric programming language that is full of capability and possible experimentation. Before I interest you in that, I'll explain to you something. What is an esoteric programming language? An esoteric language (or an esolang, colloquially), is a programming language designed to not fit the coding 'norms' or conventions. Take an example: BF ('BF' is an abbreviation and euphemism of brainf***). A standard language would notate a 'Hello, World!' program as something like: print ( " Hello, World! " ) BF, on the other hand, requires something like this: ++++++++ [ > ++++ [ > ++ > +++ > +++ > + <<<< - ] > + > + > - >> + [ < ] < - ] >> . > ---.+++++++..+++. >> . < -. < .+++.------.--------. >> +. > ++. As you can see, BF, like most esolangs, is different: it's hard to write and a puzzle. Whizz is inspired by BF, as its incrementing, decrementing and looping are inspired by it. I made Whizz because I thought languages like BF were way too monotonous to write. Esolangs should be hard and puzzling to write, but not laborious. BF requires you type '+' as many times you want to increment (without loops): so you have to find shortcuts and unscalable solutions, just to achieve your goal. In Whizz, just type that incrementation repetition count before the '+' sign, and there you have it! These wonderful features that Whizz boasts keep the challenge in esolang-ing, but contradictorily makes it more 'scalable'. Another notable feature is functions: the epitome of order. An example of a Whizz program would be: zeroToNine { [ create variables ] counter 10+ [ track state ] char 48+ [ print this one ] space 32+ [ space char ] ( char!+ [ print and increment char ] counter-; [ decrement counter and end if zero ] space! [ print space ] ) } zeroToNine* This, self explanatorily, outputs '0 1 2 3 4 5 6 7 8 9'. Again, in minimized form: c10+n48+s32+(n!+c-;s!) I genuinely hope you experiment with Whizz, and solve puzzles & challenges with it, as if it were BF! Install it

If you just want the recommendation: for the ordinary AI-generated image an inference job hands back — a 2 to 8 MB PNG — do one plain object PUT into your S3-compatible storage and stop there, because multipart upload only earns its complexity when a single artifact is big enough that losing a transfer halfway through costs you real money to redo, which for my team starts somewhere north of 100 MB. Everything below is about that threshold, and about the operations bill you pick up the moment you cross it. I run the platform roadmap for a team that renders a few hundred thousand images a month, and I count pages before I count features, so read the rest with that bias in mind. Should I use multipart upload for large AI-generated images, or a single object PUT? Multipart solves two narrow problems: a payload too awkward for one HTTP round trip, and a transfer you refuse to restart from byte zero. A 6 MB PNG has neither problem. The shape of the flow is always the same wherever you run it. You start a multipart upload and get back an upload id, you push each part under that id, you collect the returned ETag and part number for every one of them, and you send the finished list back in a complete call that stitches the object together server-side. Parts have to be at least 5 MiB on Amazon S3 and on every S3-compatible store I've tested against, with the final part exempt, which already tells you the feature was designed for objects measured in hundreds of megabytes rather than for a batch of thumbnails. Where it genuinely pays off in an image pipeline is the long tail: a 4-gigapixel tiled upscale, a nightly ZIP export of a customer's whole render history, a raw latent archive somebody in research wants kept for a year. Those are the jobs where a dropped connection at 80% is a real incident and not a shrug. For everything else, one put is one line of code and one thing to monitor. There's a second cost that people underrate, and it's the one I'd argue about in a design re

I went looking for a small improvement to an open-source tool. I found a number that pointed the wrong way, and then I found out why it had to. Live demo — paste your own file and watch it happen: https://pin-on-expand.onrender.com The setup Paritok is a 4B model that compresses AI coding-agent context. It sits between your agent and Anthropic or OpenAI, squeezes the file reads and tool output, and tells you what it saved. It's genuinely good work. Trained on 45,000 real agent trajectories, so it knows a function signature matters more than a debug line. Apache 2.0. Runs on a consumer GPU. Their benchmark numbers hold up. I wanted to build a policy improvement on top of it. To prove my improvement helped, I first had to measure what stock Paritok cost. That measurement is the whole story. Two numbers that disagree One coding-agent session. One 20,005-token file in context. Paritok's own /stats endpoint: 64.0% of input tokens saved. What the provider was actually POSTed: 69.2% more than sending the file with no compression at all. Same session. Same file. Both numbers correct. Where the missing tokens went Paritok is non-destructive by design, which is the good part. Compressed content gets tagged [REF:id] , and when the model needs the exact original it calls an injected expand_context tool to pull it back. Lossy on the wire, recoverable when it counts. The proxy answers that call itself . It appends the full original to a proxy-local thread and POSTs that thread upstream a second time. And stats is computed once, in process_request — before that loop runs. post 0: 6,919 tokens compressed request ← counted by /stats post 1: 26,924 tokens carries the full original ← never counted ───────── billed: 33,843 Then it compounds. The proxy conceals the virtual exchange from the client, so your agent never sees it. Next turn the agent re-sends the original file, Paritok re-compresses it to the same reference, and the model expands it again. Every turn. Forever. In fairness:

OpenAI has published a post-mortem examining an unusual pattern in its model testing: recurring references to “goblins” and “gremlins” in model outputs. The company’s official post, “Where the goblins came from” , published on April 29, 2026, frames the behavior as an emergent effect of reinforcement learning and human-feedback dynamics, not as a new product feature. Its practical message is more consequential than the metaphor suggests: unexpected model personas can affect the consistency, safety, and reliability that developers expect from AI systems. The published analysis provides the substantive context behind recent attention to a purported “goblin-level” post. Rather than indicating a model launch, OpenAI’s account suggests a narrower but important lesson about how optimization signals can inadvertently reinforce patterns in language models. For organizations using LLMs in production, the relevant question is not whether goblin-like language is amusing. It is whether teams can detect and address unexpected behaviors before those behaviors influence customer-facing, operational, or high-stakes workflows. What OpenAI documented OpenAI said the “goblin” and “gremlin” metaphors appeared during GPT-5.x testing and RLHF training. The company reported a notable increase in goblin-like language during GPT-5.5 testing when Codex was being evaluated. According to the post, the pattern emerged from reward-signal dynamics : persona-like responses were inadvertently reinforced through reinforcement learning and human feedback. That distinction matters. OpenAI does not characterize goblin behavior as a fixed capability or intentional model identity. It describes it as a byproduct that can arise at scale when a training and feedback process favors certain output patterns. The episode is therefore best understood as an alignment and evaluation lesson, rather than evidence of a separate “goblin” model, feature, or policy release. OpenAI also described a mitigation introduced

What Is Temperature in AI? (And How to Stop Getting Poetry When You Asked for a Grocery List) Remember Magic 8-Balls? Those plastic oracles you'd shake for life advice, only to get "Reply hazy, try again" when you asked if your crush liked you back? Imagine someone added a little dial on the bottom. Turn it all the way to zero and the thing becomes painfully predictable, only ever offering "Yes" or "Most likely." Crank it all the way up and suddenly it's inventing answers that never appeared in the original twenty options, things like "Ask your neighbor's cat" and "The moon suggests Thursday." That dial is temperature, and every AI language model has one. How the dial works Temperature is a setting, usually ranging from 0 to 2, that tells an AI model how much risk to take when picking the next word. The model calculates the probability of every possible next word, then has to pick one. At low temperatures, it plays it safe and picks the most probable option almost every time. At high temperatures, it's willing to gamble on unlikely choices further down the list. This is why you can ask ChatGPT the exact same question twice and get a straightforward answer on Monday and what appears to be surrealist fiction on Tuesday. When you ask ChatGPT to write a professional email at temperature zero, you'll get "Dear Sir or Madam, I am writing to follow up on our previous correspondence..." every single time you hit enter. Set temperature to 1.5 and it might open with "Greetings, fellow traveler of the inbox wilderness" because that phrasing, while statistically improbable, is now in play. Why boring is sometimes good At temperature zero, you get the most boring dinner guest imaginable. It always picks the single most likely next token (the technical term for a chunk of text, usually a word or part of one). No variety, no surprises, just the statistical favorite every single time. This turns out to be perfect when you need factual accuracy, code that actually compiles, or data

There is a number going around that roughly half of all remote MCP servers are dead. I had repeated it myself, in the README of a tool I published. I could not find where it came from, so I measured it. The answer is that about one in ten is actually broken. The "half" figure appears to come from counting servers that require an API key as if they were down. Here is the method and the full breakdown. What I measured On 29 July 2026 I pulled every entry from the official MCP registry — 1,200 servers. Of those, 297 had status: active and advertised a remote endpoint URL (the rest are stdio/local packages with nothing to probe over the network). Each got one anonymous JSON-RPC initialize over streamable HTTP, with a 10 second timeout: { "jsonrpc" : "2.0" , "id" : 1 , "method" : "initialize" , "params" : { "protocolVersion" : "2025-06-18" , "capabilities" : {}, "clientInfo" : { "name" : "mcp-uptime" , "version" : "0.1.0" } } } Then I classified the response: a valid result containing protocolVersion or serverInfo is up, 401/403 is auth-gated, and everything else got bucketed by its actual failure. Results (n = 297) Result Count Share Completed an MCP handshake 133 44.8% Auth-gated (401/403) 134 45.1% DNS failure 8 2.7% Server error (5xx) 6 2.0% Not found (404/410) 5 1.7% Redirect (307/308) 4 1.3% Timeout 2 0.7% Non-MCP response 2 0.7% Other (400, 405, connection) 3 1.0% Reachable: 267 (89.9%). Genuinely broken: 30 (10.1%). Where "half are dead" comes from Look at the first two rows. 55.2% of these endpoints will not complete an anonymous handshake — and that is suspiciously close to the number people quote. But 134 of those 164 are returning a clean 401 or 403. They are running. They are answering. They want an API key, which is a completely reasonable thing for a hosted service to want. Counting those as dead inflates the failure rate by roughly five times. This matters beyond pedantry: if you believe half the ecosystem is rubble, you build defensively against the wron

Most "beginner" Python courses aren't actually beginner courses. Lesson one usually opens with variables or print(), quietly assuming you already know what a terminal is, how to install something, or what a .py file even means. That assumption is exactly where most self-taught learners bounce — not because Python is hard, but because the ten minutes of orientation that would've made everything after it make sense got skipped. I built Codes Are Simple to start there instead. Session 1, Level 1: what is code, where do you type it, how do you open Command Prompt, how do you install Python and verify it worked. Nothing assumed. What it actually is A 45-session, self-paced curriculum — Python: The Universal Language, Zero to Pro — split into 10 tiers, from absolute foundations through OOP, files/errors, practical CLI projects, web/APIs, databases, and a final professional capstone. It's the first of a planned multi-track catalog on the same platform (web dev, AI, cyber, and data are mapped and coming next). Every lesson — all of them, across all 45 sessions — follows the same repeating shape: explanation → main example → 2 extra examples (variations/edge cases) → common mistake (shown alongside its fix) → practice → extra practice That "common mistake" section is the part I actually care about most. Almost every course I looked at shows only the correct code. This one shows what actually breaks for a beginner and why — because for someone learning alone with no instructor in the room, the error message is usually where the real learning happens, not the clean solution. Checkpoints land at fixed points mid-session (after Level 2, Level 6, and Level 9 — not just at the end), and every session closes with a real capstone project, not a toy exercise. The stack Cloudflare Workers + Pages + R2 — the site and all curriculum content, served as versioned JSON per session GitHub — version control for the whole content pipeline That last point is the part I think this community wil

Exploits can give persistent server access that survives credential rotation and disk re-imaging.

Apple's iPhone and Mac sales are on the rise even as a global memory shortage squeezes device makers. In its third-quarter earnings report released on Thursday, Apple revealed that iPhone sales jumped 22 percent to $54.25 billion and Mac sales increased 29 percent to $10.35 billion, driving the company's overall revenue to $109.4 billion. The […]

"Chilling message to all broadcasters: carry speech we don’t like at your peril."

I am not by any means an expert at finance but I think I do now have some advice for people who are: Do not name your hedge fund anything that will be hilarious if it blows up. Don't use a name like "Long-Term Capital Management," or "Amaranth Advisors" (named for the floral symbol for […]