Electron Soccer (2026) by kimchipenguin2 for Atari ST.
A two-player port of the C64 oddity where you steer particles, not players: bump the white ball to pass on your speed, scoring top or bottom. ⚽ Now on ST low and high res 🕹️
Search
1000 results for “Not_AI”
-
FYI: Draft bill tied to Anthropic would let courts halt AI models for 45 days: Federal draft last edited September 29 sets a 10^25 compute line and a $500M revenue test. Which state frontier AI rules would survive it, and which would not? https://ppc.land/draft-bill-tied-to-anthropic-would-let-courts-halt-ai-models-for-45-days/ #AIBill #ArtificialIntelligence #TechRegulation #AIModels #LegalTech
-
What makes an agent genuinely useful? Guardrails. The Guardrails That Make Agents Useful considers why clear boundaries matter when agents are asked to act, not just respond.
👉 https://zalt.me/blog/2026/10/guardrails-make-agents-useful
-
Disclosure: I run reaguide.online, an independent, lab-verified tutorial site for REA. It is not...#reverseengineering #aiagents #security #tutorial #software #coding #development #engineering #inclusive #community
REA vs Ghidra: I Ran Both in a Lab for a Week - Here's the Honest Comparison -
High-end AI chips sit mostly in data centers owned or leased by a handful of hyperscalers, and in 2026 that capacity got more expensive, not less. https://hackernoon.com/the-decentralized-gpu-marketplace-breaking-the-big-tech-compute-monopoly #ai
-
An auditor has to review what an AI did. The cheapest way to let them do that is to hand them an...#ai #security #agents #software #coding #development #engineering #inclusive #community
An Auditor Is Not an Admin: Giving Someone the Evidence Is Not Giving Them the System -
An AI coding agent’s overall accuracy score can hide where it helps—and where it creates more review work. Evaluate agents by task category, define acceptance criteria in advance, and track outcomes beyond whether a pull request is accepted.
A study of 7,156 pull requests offers a useful example of why this matters. The paper Comparing AI Coding Agents: A Task-Stratified Analysis of Pull Request Acceptance compared five agents across nine task categories over 32 weeks.
Why task category belongs in the evaluation
The study reported an 82.1% acceptance rate for documentation tasks and 66.1% for new-feature tasks—a difference of about 16 percentage points. That gap was larger than the typical difference between agents in many categories.A single blended score can therefore obscure important differences. An agent that is useful for documentation may not be the right choice for feature work, where requirements interpretation, dependencies, and testing create more ways for a change to miss its intended behavior.
Acceptance rate is a useful outcome, but it does not measure code quality, security, maintainability, or developer productivity on its own. Treat it as one signal in a broader evaluation.
What the study found about individual agents
The results did not identify one universal winner. OpenAI Codex had consistently strong results across all nine categories, with acceptance rates ranging from 59.6% to 88.6%; the paper also reported statistically significant advantages over other agents in several categories.Category leaders varied: Claude Code led documentation at 92.3% and performed best on feature tasks in the reported results, while Cursor led the fix category at 80.4%. Devin was the only agent with a consistent positive acceptance trend during the 32-week period, increasing by 0.77% per week. The other agents were described as largely stable.
These are historical results from one dataset and time window. Use them to shape questions for your own evaluation, not as a substitute for testing against your repositories and review standards.
Build a task-stratified evaluation
1. Choose categories that reflect your work
Separate evaluation sets for documentation, bug fixes, feature work, refactoring, testing, and other relevant tasks. Start with the categories that make up your intended use case. Keep results separate by category rather than combining every pull request into one score.2. Make tasks assessable
For each task, state the objective, boundaries, relevant files or services, expected behavior, and validation requirements. Keep requirements connected to the implementation and test evidence so reviewers can judge whether the change meets the original intent.Feature work warrants particular care: the study’s 66.1% acceptance rate for new features was lower than the documentation result. Clear requirements, linked work items, assigned review ownership, and test evidence help make these changes easier to assess. Provide a way to revise rejected changes rather than treating rejection as the end of the process.
3. Track more than acceptance
Record acceptance rate by task category, but also capture review time, rework, escaped defects, test failures, and how much human editing was needed before acceptance. These measures help distinguish an accepted change that required substantial correction from one that was ready with little revision.4. Preserve evaluation context
Record the agent and version, task category, repository context, and acceptance criteria for each evaluation. Repeat the benchmark periodically: models, prompts, repository conventions, and review practices can change, so results may not remain comparable without this context.Turn results into delivery controls
- Classify the intended work. Decide which task categories are in scope and review results separately for each.
- Set quality gates. Specify required tests, reviewers, security checks, and documentation before expanding agent use.
- Keep work traceable. Link requests, requirements, implementation, reviews, and outcomes so the team can investigate failures.
- Compare total effort. Consider review time, revisions, rejected pull requests, and post-release issues alongside acceptance.
- Reassess regularly. Repeat measurements when agents, repositories, or internal practices change.
The practical takeaway is to select and govern coding agents at the task level. The study’s category differences show why a broad ranking can mislead; a local benchmark, explicit acceptance criteria, and ongoing measurement give teams a more useful basis for deciding where an agent fits.#ai #devtools #softwareengineering #testing #software #coding #development #engineering #inclusive #community
Evaluate AI Coding Agents by Task, Not by One Score -
AI pahoehoe lava with some random "volcano" in the background just gives me pain in the heart region. it's just so wrong. this is NOT what pahoehoe looks like. it's not just a bunch of bumps on the ground. pahoehoe lava reflects how lava flowed across the land and where it settled.
it's almost offensive. by which i mean, i'm kind of offended. hawai'i is the kind of place where people really KNOW what it looks like when you go to a particular place and they value when something looks correct. because otherwise you're not honoring the land...you're disrespecting it, right?
for a contrast, check out the beautiful mural in pic 2 - a mural on a local middle school. this is what it looks like to honor the land and to honor the actual plants and animals from here. this is socially useful because kids who walk past it every day will know what their local plants and animals look like. it's got nene. it's got pueo (nesting on the ground - correctly). it has hapu'u tree fern, it has the right birds in the trees. it has the right orchids. it has the right mauna kea in the distance. it may not be an actual valley, but it has all the elements that an actual valley would have. plus, it's beautiful! its beauty honors the beauty of the land.
(AI image is from https://ainaforce.com/blog )
-
"Moritz Hardt, author of a new book on benchmarks, believes these 'standardized tests' for artificial intelligence are fundamentally broken — but not unfixable."
"He Wants AI Benchmarking To Be a Science, Not a Dark Art"
https://www.quantamagazine.org/he-wants-ai-benchmarking-to-be-a-science-not-a-dark-art-20261016/
-
Traffic from ChatGPT and Perplexity often converts better than ordinary organic search because the buyer arrives with a recommendation.#seo #ai #saas #analytics #software #coding #development #engineering #inclusive #community
AI Referral Traffic Converts Better. That Does Not Mean You Should Chase It Blindly. -
AAIF added four projects on 22 October 2026. Goose was already there. Here is what that changes for a Flisk laptop stack, and what it does not.#flisk #goose #docker #ai #software #coding #development #engineering #inclusive #community
Today's AAIF Announcement Does Not Add Goose to flisk.sh -
Disclosure: Independent guide from Fuxux, which is one of the tools compared. We are not affiliated...#ai #mcp #chatgpt #productivity #software #coding #development #engineering #inclusive #community
Social Media Schedulers That Work in ChatGPT and Claude (2026) -
FCC chair warns media not to run the “fake polls” Trump keeps complaining about
Ex-FCC chairs say Brendan Carr aims to scare broadcasters into self-censorship.
#news #politics #uspol #uspolitics #GOPCult #trump #ethics #corruption #maga
-
"Brain Drain Is a Bigger Threat to National Cybersecurity Than AI
What good is investing in offense if you’ve hobbled your own defense?
Whether we’d admit it or not, most of us by now have offloaded some cognitive effort onto AI. Chatbots and agents have become so commonplace that it’s often tricky to avoid them. The stakes of this use are usually pretty low; we may compromise some of our own critical thinking skills and creativity in the process, but the harm stops there. In some cases, however, an over-reliance on automation can have major downstream consequences.
Cybersecurity is the most notable example. Thankfully, that’s a field that most of us almost never have to think about, because most of the time, the computer systems running critical infrastructure—from water treatment facilities to the stock market—function as they’re supposed to. But many of those systems haven’t been updated in decades. In July of last year, a report from the U.S. Government Accountability Office found that the Department of Defense’s in-house IT system has been in use since before the first Apollo moon landing. (The Department of the Treasury and the Environmental Protection Agency weren’t far behind.) 'These agencies’ missions are essential to government operations such as health care, critical infrastructure, tax processing, and national security, and these legacy systems provide vital support to the agencies’ missions,' the GAO wrote in its report. COBOL, a computer programming language invented in 1959, is still widely used on IBM mainframe computers across government agencies, so much so that during the pandemic the governor of New Jersey was forced to put out a call for 'cobalt' (his word) programmers to help maintain the state’s decrepit and overwhelmed unemployment system.
Unfortunately for the cybersecurity of the United States, much has changed since those 'legacy' cybersecurity systems were first installed. Programmers today—perhaps more than any other profession—rely on AI in their day-to-day work. And while that may have led to some efficiency gains, it’s also turned much of modern programming itself into a black box. In a recent study with over 50 (mostly junior) software engineers, Anthropic found that using AI for coding 'led to a statistically significant decrease in mastery.' And in a May survey of more than 850 cybersecurity professionals, two-thirds of them said they were now spending more of their time on the job reviewing AI outputs and trying to decide whether to trust them or not. Much of modern cybersecurity, in other words, revolves around just trying to interpret AI."
https://gizmodo.com/brain-drain-is-a-bigger-threat-to-national-cybersecurity-than-ai-2000778508
-
World-first discovery harnesses mRNA in the search for an HIV cure - Doherty Institute
"Australian researchers have made a major breakthrough in HIV research by repurposing the same mRNA delivery system used in COVID-19 vaccines, not to prevent infection, but as a potential strategy to find a cure."
https://www.doherty.edu.au/articles/world-first-discovery-harnesses-mrna-in-search-for-hiv-cure/
#aids #health #medicine #mrna -
'Sources on credit desks told CNBC that the high-yield debt market is effectively on hold until after the midterm elections in two weeks given ongoing concerns about the political pushback that data centers are facing. One source, who asked not to be named because of the sensitivity of the matter, said fixed income investors “want permitting and regulatory requirements to be substantially resolved” before they commit financing.'
https://www.cnbc.com/2026/10/21/coreweaves-market-confidence-erodes-as-bond-yields-spike-stock-sinks.htmlHahahahaha. 😂
-
AI may eventually kill us all, but first it will make us miserable.
You have probably heard of the security weakness in the PHP programming language. If not, it is urgent that you become informed on this issue. It affects a massive number of websites and internal applications. Hackers can bypass security credentials and get full access to servers. Patches are available and should be applied as soon as possible.
Because we have entered a heavy activity period before the December holidays, many companies are in lock-down mode. No one wants to risk upgrades during the most busy season of the year. But this is an exception. A five-alarm fire, if you will. The patches are easier than the Year-2000 event, but much more critical to get done now.
Best wishes to my IT colleagues, and my sympathies.
-
-
Adding AI to a product is easy. Adding it where it actually helps is the work.
We build Claude and OpenAI integrations into web and mobile products: a feature your users reach for because it saves them time, designed and tested like the rest of the product, not bolted on for an announcement.
Book a free consultation at codrlabs.com
-
I thought #AI was supposed to cure cancer, fix the climate crisis, and free us from labor so we can spend our time on art and science and philosophy.
That does not seem to be what's happening.
-
https://www.europesays.com/ie/746167/ Nvidia reveals plans for ‘an AI factory in a shipping container’ to build data centers in weeks, not months #401(k) #AI #ArtificialIntelligence #ArtificialIntelligence #Business #BusinessNews #Economy #Éire #Finance #FinancialInformation #IE #Investing #InvestmentTools #investor #Ireland #MarketNews #mortgage #MutualFunds #PersonalFinance #quote #RealEstate #Retirement #stock #StockResearch #StockValuation #stocks #SuzeOrman #tax #Technology #TrackPortfolio
-
A New York appeals court asked where the cases came from. The lawyer couldn’t say which AI tool he used.
The check failed, and then so did the answer to where the work came from. Both rested on one person remembering. The fix is a process that does not.https://aifueledculture.wordpress.com/2026/10/22/the-court-asked-where-the-cases-came-from/
-
On-premise AI inference beats cloud APIs above a break-even point set by utilisation and request volume, not by which slogan you prefer.
https://mickai.co.uk/articles/crossover-point-on-premise-cloud-inference
-
Because they're insisting on this unfettered #ai-everything nonsense, going Ꝭull '#digital money' will eventuαlly Ნecome an impossibility.😑 It was αlreaᲫy an infeasible goal bc of infrαstructure and the Ꝭrequency of nαtural Ძisαsters.😑 Not only that, ppl behind it are humans. Imagine the diabolical shenanigans.
#tech -
Outlook signature set up on your work PC, missing when you open Outlook in a browser? Not a bug.
Microsoft: signatures "are stored locally on the device where they were created and may not be available when you use Outlook on another device."
And if Outlook's menus don't match the guide you're reading, you're probably in the other Outlook. New Outlook: Settings › Accounts › Signatures. Classic: Message › Signature › Signatures.
-
Not a good day to be able to read, the newspaper in this case.
- Dutch politicians “caught” using genAI for their speeches and arguments
In the same newspaper:
- earth will warm even more than we already thought, models predict. (Not a single mention of data centers/AI in this article)Idk man, what the actual fuck are we doing.
There’s at least one Dutch politician (Laurens Dassen) who is openly an AI skeptic and he’s being ridiculed for it.
What the actual fuck is going on???
-
AI's shaping election messages more than political ads. It's a wild ride, but at least it's not ads. — Spectrum News https://gigcitygeek.com/go/DzNeDjcv
-
My day job landed on this rule about AI usage in documents and code. Instead of policing people what they use and how, and try to delineate what is AI and what is not, it's basically like this:
you're free to use whatever you want to do your job, but it's your name on the outcomes, and if they suck that's on you.
and so, when people produce AI generated documents that are full of shit, the artefact goes against their own performance.
which i think is fair.
-
AI agent runs often finish before queues, workers, webhooks, and downstream systems reveal the result. Trace the outcome, not only the request. https://hackernoon.com/your-agent-trace-ends-before-the-outcome-exists #aiagents
-
AI tools write answers that sound sure, including the wrong ones. You do not need to be technical to catch many of them. Ask these five questions, in this order, in the same chat.
- "What would have to be true for this to be wrong?" A good answer names real conditions. A vague one tells you the first answer was weaker than it sounded.
- "Which parts are you sure about, and which parts are you guessing?" Mark the guesses and check only those.
- "What did you assume about my situation?" Many bad answers are good answers to a question you did not ask. Your country, your budget and your skill level are the usual missing pieces.
- "Give me one source I can open myself, and tell me if you are not certain it exists." Then open it. If the link or the title is not real, treat the rest of the answer with care.
- "How could I check this in five minutes without you?" A search, a calculator, a phone call, a look at the official page. Do that one thing before you act on the answer.
None of this makes the tool perfect. You stay the one who decides.
We are TIC, a small studio that makes plain-language AI guides (ticassociation.com). The list above is free to copy and share.
Drafted with AI help and approved by a person before posting.#ai #beginners #productivity #software #coding #development #engineering #inclusive #community
Five questions to ask before you use an AI answer