Todays hiring technology is experiencing an arms racebetween hiring managers and candidateswith both groups attempting to process the most applications with the least human contact: the result is a bizarre mockery of the idea of hiring, absent humandecency.Having recently changed my day job, I have had a lot of contact with the tools and practices of hiring: the majority of the aforementioned, alongside the treatment of candidates, has been appalling. I cant claim to point to the ultimate cause, but the result is the dehumanization of candidates by analytical, AI and automated solutions that hiring managers turn to in order to deal with the unmanageable deluge of resumes from candidates who have turned also to scale and toAI.I will address the following irritants:PDF as a prime means of data transmission (forresumes)Applicant Tracking Systems (ATS) and the new need to keyword-optimize ourresumesThe habit of companies to decline candidates without explaining their reasons or giving candidates a means toreplyUnsurprisingly, I have a few recommendations forreform.PDFs and dataexchangeMore or less, all jobs posts today require you to detail your work history, skills and education. Most commonly, candidates can have the hiring system read their resume, or have the data pulled from their LinkedIn profile, or in more rare cases submit it manually.All of these options are bad. Manually copying and pasting ones experience into the requisite boxes is slow and tedious. LinkedIn is fast and direct, but it is a monopoly in its space, which is a problem initself.Then we have the PDF-parsers, which range from tolerable to hilariously bad: garbling dates, reversing fields like locations and title, and sometimes apparently plucking information from nowhere. This is even after formatting my resume within special tools designed to create parsable PDFs. Note that for jobs that dont allow you to submit experience via LinkedIn, often you must submit either by PDF parse or manually. This, compounded over days and weeks of applications, is humiliating.In addition, PDFs have innate problems as means of information transmission. Most of the PDF standard has been in the public domain since 2008, but working with PDFs feels like working with a proprietary standard like.doc: one must wrestle either with free/open tools that are at best fair quality or use full-featured but bloated and glacially slow proprietary tools, and work around inexplicable differences in formatting betweensystems.With PDF, something is alwaysbroken.This of course raises the question: why do we use PDFs? We do because PDFs are a means of fixing text and images on a page with purported portability between systems. But why does it actually need to be on a page? I say it doesnt: for the majority of candidates, what matters is the organized textual expression of what they can do. Assuming the text is readable, the layout is irrelevant except as a showcase of design skills for relevant professions.When creating a PDF with a tool designed for prettiness or parsability, one is constantly fighting the word-countIndeed, layout is a stumbling block: the best-looking PDFs that can be made by non-designers are built in PDF-building tools. These tools are decent, but the prettiness comes at the cost of inflexibility: no tweaks to layout or font size are possible, meaning that one is constantly fighting character count, for example, to keep things on one page. PDF, like email, comes also with the anachronism of baked-in line-breaks.This text, copied from the PDF in the above screenshot, has carriage returns inserted not by the writer but in order to format thePDF.The text, shown without the offending line breaks. (The width does make for more difficult reading, but this is properly solved by adjusting column width, not inserting carriage-returns.)Today text, just text, verges on universality and openness, flows to fit the allotted space offered by various devices and applications: not so withPDF.A moment spent on deleting spurious carriage returns or editing to keep things on a single page is a moment spent on something irrelevant to what actually matters: ones aptitude.ATSs and keyword optimizationTheres good reason to think that candidates who optimize their resumes with keywords found in the job description have a better chance of being interviewed. This is supposedly because ATSs perform a keyword analysis on resumes as a purported measure of relevance. As with search engine (SEO) keyword optimization, this is a signal that the system is broken: the act of taking words from the job description and putting them into your resume is formulaic, therefore demonstrating nothing nontrivial about ones qualities.The fact that its possible for it to change ones chance of getting an interview suggests that hiring managers are dealing with a scale of applicants and/or lack necessary tools such that they cannot make real judgments on candidates.A screenshot from JobScan, a tool that scans the job description and your resume and recommends keywords to insert. Taken from JobScans tutorial.Meanwhile, if you eliminate candidates based on keyword analysis, this will eliminate candidates who have either not heard that keyword optimization can help or who refuse to do so, feeling that it is deceptive (these people should behired).The picture of resumes will therefore, become something like the Web in around 2010: having discovered keyword stuffing, webmasters gained rankings for thin, derivative rewrites of rewrites, while real ideas languished.As with SEO, this distortion comes from a bad information system (the Web) that lacks a decent indexing, categorization and intercomparison system. Thus, instead of using resumes to exchange information, we have a new arms race: the employers use ATSs to filter us, and we keyword optimize: this arms race cannot be won, and we will continue it at the expense of our time and self-respect.To see the evidence against me, and to confront myaccuserId like to draw attention to an excellent presentation by Casey Muratory, given earlier this year during FUTOs Dont Be Evil Summit. Muratory taks about the process of people being thrown off platforms such as Twitter and YouTube, and claims that these processes would be fairer and have better outcomes if the platforms in question implemented some norms from the legal system, notably the right to know the evidence against oneself and to discuss the situation with the person making the decision.*https://medium.com/media/99d29bcac2ad4f66f918ab339614238c/hrefI think that we should apply something similar to hiring. In hiring, both of these principles are violated almost universally, see an example application responsebelow:From: no-reply@us.greenhouse-mail.ioSubject: An update on your application to *****Partner Marketing Manager*****Dear Oliver,Thank you for your application to the Partner Marketing Manager***** role at *****. We are writing to let you know that we have reviewed your application. We regret to inform you that we shall not be progressing your application further for this particular role.We appreciate that you considered us for your next career step and hope that you maintain interest in ***** and our products. If you would like to apply for other new openings we would be happy to consideryou.Regards,*****There is no explanation of how my application was lacking. Too little experience? To much of a jump from my current title? Prefer someone on the West Coast? Not enough keywords?How did they make the decision? AI? A human? An algorithm?In that they can arrange to have this email sent and to include the job title and my nameautomaticallythey could include their reasoning. Not including information that could be included at no cost to anyone is called hiding it. I suspect they hide it because they dismissed my application out of hand algorithmically or with the help of AI: its their right to do so, but I have the right toknow.How a database of hiring decisions mightlook.This would of course be galling news, but it would help us candidates immensely to who (or what) and makes these decisions, and how. Indeed, knowing that an algorithm made the decision, say, and that it was based on lack of requisite keywords would help me know that this is something I should focus on in future applications. And if they admit to using AI for this purpose, thats the evidence we need to pushback.Lets say, dear reader, that you disagree that we candidates have a right derived from common decency to this information: fair enough. But let me put it this way: informing candidates about how decisions are actually made will benefit the companies, in that our applications will be better suited to their systems. And, if we know that a given application is destined to fail, we wont waste their time withit.Of course, truly frivolous applicants dont have such rights, but Im treated like a frivolous applicant much of the time, even in response to heartfelt cover letters, detailed answers, even after for one application looking at the % of people in the UK who got the same grade as I did in English to see how good, relatively, my gradewas.It would be useful to know what the hiring manager or machine actually did. Did they look at my answers to their questions? Watch the introductory video Imade?I once experimented with taking time to write long, detailed cover letters that respond to the companys situation and brand, and the results were the same as for any other application I had submitted. Did someone read it? Tell us what we need to do and what will help ussucceed.We will doit.Then theres the no-reply emails. If the company tells me its a no from a no-reply, I have no recourse to dispute the decisions and no way to ask for feedback, or, so to speak, to confront my accuser. This is a sad return of the idea that a computer (which, lets face it, is likely making most of these decisions) cannot err, and by extension that there is no need to be able to query its decision. I can of course fill in the companys contact form: but what good is that actually going todo.Remember, mistakes happen: it may seem absurd, but what if I was the best candidate and the hiring manager accidentally pressed the wrong button? AI makes frequent and colorful mistakes. What if I made a mistake and, realizing it, wanted to let them know after my application had been rejected. Often Im confronted with US applications that require a GPA. I dont have a GPA because I studied in the UK: will they read my explanation? How can I be sure if I cantask?ProposalsBelow, I set out some proposals on how to improvethings.A better information systemThe solution to the absurdity of PDFs is obvious: job seekers should maintain their work history, skills and interests in a standardized format, rendered in plain text. Ideally it would be accessible online via a URL, but you could equally store it as a file. Note that this is not a document; the fields, like employment dates, titles, etc. would be stored to allow universally accurate parsing by computer.Designers and others for whom the visuals are important could maintain both designed resumes and a system like this to optimize for both sides of the equation.My company, HSM, is building a broader solution to the superset of this problem: the proper organization of text, its ownership and control byusers.No no-reason, nono-replyNo-reason decline emails can be abolished immediately at little to no cost: there is no excuse. No-reply emails should be abolished also: the initial result of this will be a deluge of correspondence (much of it warranted, Im sure) but then hiring managers will be forced to raise the bar to reduce the volume of applications: this is good thing, we candidates apply to too manyjobs.No AIFor the most part, I think that AI (read LLMs) should have no application in hiring (and probably most other fields). Using it abdicates responsibility to a system whose decisions are definitionally impossible to interrogate, which is immoral. For more on this, see my longer discussion of AI, Artificial Intelligence: The Soul of Soulless Conditions.If you dont have time to read it, I defer to a famous slide from an IBM presentation on thesubject:Talk to failed candidatesA friend of mine, when laid off, asked whether there would be an exit interview. His boss responded by saying that there would not be, as his was an involuntary termination. What a odd proposition: that this individual, because he was being let go, had nothing useful to offer the company by way of feedback or praise. The same is true for failed candidates: they have a lot to say of much use, but nobody reachesout.I got a survey once; the first question was: Did the job description make it clear to you what the role would entail? I didnt read anyfurther.Walk thewalkFinally, any company that uses PDF parsing, keyword-oriented ATS systems, AI, and that hits candidates with no-reason, no-reply messages should mandate that any hiring manager and anyone who has a hiring manager report to them must apply for their own job with such a system every quarter. This will I hope make them realize firsthand that these systems would be hilarious if they werent so dehumanizing, and will spur change from a sense ofdisgust.ConclusionThere is foul play on both sides here: candidates apply to hundreds of jobs when they probably shouldnt, and use AI to help when they definitely shouldnt. This overwhelms hiring managers, who then need systems to deal with the quantity. The question of who started matters less than the fact that both sides are stuck in this trough: they use computers on us, so we act like computers in order tosurvive.A candidate who applies for a reasonable number of jobs or shuns keyword optimization software hurts only themself; companies must act first, as they have the scale and the clout to do something.All this seems to have happened thanks to the computer adding scale to and subtracting personality from our interactions. Indeed, interacting with real people can be hard and awkward, especially if its bad news. But surely personality is what its all about. Indeed it feels that we subject our personalities to such scale that they risk thinning out into an undifferentiated haze or, like a balloon in a vacuum, expanding rapidly and goingpop.*I note that the question of online platform access is controversial and polarized. However, if you feel that this cause belongs to people who dont think like you do, I encourage you to seek examples of de-platforming when someone like you was the victim, or even to read about non-political examples (all exist). I ask you then to tell me if Muratorys proposals would make things worse: please comment or contact me. This is for now putting the costs of implementing his proposals to one side: they would be considerable, but the most relevant platform companies have similarly considerable amounts ofmoney.Why is hiring software so impersonal? was originally published in UX Collective on Medium, where people are continuing the conversation by highlighting and responding to this story.

Smart devices, on their own, are nifty, but one-note. You can turn things on or off, adjust volume or brightness, or tell your vacuum where to go. But automations are what make smart homes awesome: you network your devices together, create dependencies, and set up if/this, then/that scenarios to automatically execute tasks. Building these automations always takes a little finesse, but now, Google's Gemini AI offers to set up automations for you. When I tested this new feature, I hoped it was for the layman who might be intimidated by the basic automation creation flow. What I found, instead, was a feature intended for advanced users who want to dive into using a new tool called the the script editor (Google's tool for writing custom automations). Most users won't have automations that require more than the standard home automation wizard that has always existed, meaning they won't need to crack open the script editor or get Gemini's help herewhich is great, because the feature doesn't seem fully baked yet. These features could become useful in the future, as complicated smart home automations become more powerful and necessary, but for now, I found Gemini unable to offer error-free assistance. Google Home, Gemini, and the script editorThere are three key components to building automations through Google services. First, there's Google Home. This is a home base for all your smart home devices, the status and settings of your wifi, and best of all, automations. If you want to turn on a lightbulb, you can do that by tapping a button or asking Google Assistant. But if you don't want to ask, automations will turn the light on based on criteria you set up, like a time of day, activity of another device, or a certain phrase you tell Assistant. You simply have to create an automation in Google Home telling it to do so, and there's a straightforward setup wizard here to help you do so. Google Home recently added a desktop app, which is still in beta, but you can create automations and control your smart home from your computer, as well. Gemini, on the other hand, is Google's large language model and AI chatbot. The company has been integrating Gemini into all sorts of Google functionality, including a chatbot available on iPhone and Android. Right now, it's most useful as an assistant that opens in the sidebar of most Google apps and can respond to real language questions. It can already do this in Gmail, finding content or tightening up the language you use, and Google Meet, taking meeting notes on your behalf. Theoretically, using Gemini to create automations for users would be a great feature, but it turns out it is actually more complicated to use Gemini than to use the current wizard. Most Google products, including Sheets and Docs have had a script editor called Apps script for a while, but it's not used often. It's a feature targeted mostly for advanced users that allows you to create your own scripts, or paste scripts other people have written to do things your products don't do out of the box. The new script editor for Google Home is similar. Although it uses a different language (YAML) than Apps script, the purpose is the same: allow advanced users to do more than the automation wizard allows. How to set up automations in Google HomeThe current user flow to build automations in Google Home is straightforward for simple tasks, but any tasks that eclipse what the current Automations wizard will do, could, in theory, require the script editor. (For the record, I've been using Google Home prolifically for five years and never once needed to use the script editor for even one of my wacky automation ideas). If you're not a coder, the script editor in any Google product can be intimidating, so having Gemini handle this for you is a big lift, if it worked. But that wasn't my experience. To get Gemini's help, you'll need to sign up for servicesNow, some notes. This does, in fact, require that you give Google permission to activate Gemini on your phone. I've been avoiding this since I activated my new Pixel a few weeks ago, but for the purpose of this article, I decided it was time to dive in. While endlessly useful, Gemini isn't free in some cases. You'll also need to join the app's public preview to access the script editor. This was a few simple steps. Open the Google Home app, go to Settings, and then click on Join Public Preview. Go to settings, and then activate public preview Credit: Amanda Blum Accessing Gemini is easier on the desktop appGoogle calls this assistive lift from Gemini "Help me script." To be abundantly clear, using the script editor on any day is not for the weak. Its a code language, YAML, is technically easier to write than other code bases like Php, Perl, Ruby, or Javascript. However, if you don't know YAML, there's a learning curve, and it is still code. Personally, I'd expect that if you pay for this Gemini assistance, you wouldn't need to know YAML. That is not the case, I found. Once you've signed up for preview, and Gemini is activated, you can open the script editor on your Google Home app or the web interface. How to access the script editor on the Google Home app Credit: Amanda Blum This was my first time using the web interface, but since there's copying and pasting involved, I'd suggest using it over the app. Plus, Gemini didn't want to pop up on the app for me, so all the more reason to use the desktop interface. On the right hand side of the script editor, you can see the Gemini prompt Credit: Amanda Blum Most scripts by Gemini had errorsTo start, I asked for help with simple automations that could be done in the wizard: tell me when it's sunset every day. This automation doesn't involve any devices, simply an out-of-the-box piece of data that Google already has. ("What time is sunset?") Gemini spit out the code, and instructed me to copy and paste it into the script editor and validate it, but warned it already had an error. The script editor confirmed this as well, and identified the problematic line of code. A mouseover explained why I was seeing the error so I could correct it. This happened over and over, where I would ask Gemini to create a script for something relatively easy ("turn on backyard lights when backyard camera detects activity") and I'd get errors. Sometimes, the script editor could explain them, sometimes it could not. But Gemini, an AI chatbot designed to answer questions from users, had no function to help me solve the problem. Some mild digging on the internet was unhelpful, since each kind of device has different possible commands, and everything I read suggested the commands Gemini wrote should work. Remember, I hadn't asked Gemini to do anything difficult yet, all the automations I asked for would have been very easy to do in the regular automation panel, without the script editor or Gemini. Errors are shown as rollover tips in the script editor Credit: Amanda Blum If you're successful with the script editor (I was able to clear some errors with a little help from the tool tips and some research), you save the script, and then it appears in your list of automations. You can turn it on and off at will, and when it is activated, you'll see it in automation history. Who is the script editor and "Help Me Script" for?As I said, I'm an automation enthusiast, and I have a ton of them running. I also code. Despite both of those facts, I maintain that most people won't need the script editor, but there are some examples where it would make sense. Anything with a double dependency would use the script editor. For instance, if I want a fan to turn on when it's over 80 degrees, I can do that in Google Home. I can also have the fan turn on if it's dark out. What I can't do is require both circumstances, unless I'm using the script editor. Additionally, the script editor would let me set conditionals, like, "if it's over 80 degrees, turn on the fan, but if it's under 60, turn on the heater." You can hack these situations together using many automations that aren't connected in Google Home, but the script editor would be a more efficient way to do so in one clean user flow. Unfortunately, to make that happen, you're still going to need to learn YAML for now, because Gemini's help isn't going to do your homework for you: For the most part, and for most people, Gemini just isn't ready to script your automations for you.

If you're tuned in to digital culture news at all, it's been impossible to miss the chatter about Bluesky over the past month. The fledgling social media sitewhich looks and acts a lot like Twitter did a decade ago, long before Elon Musk bought it and renamed it Xis exploding in popularity, doubling in size over the past two months to around 20 million users.Suddenly a site you'd probably only heard of if you spend a lot of time thinking about your social media diet is topping the App Store charts and attracting a flood of folks looking to exit X as it has drifted toward promoting toxicity and even disinformation in the lead up to, and aftermath, of the U.S. presidential election. All of the reasons for the site's recent boom aside, if you're a new user, the most important thing to you is probably figuring out how to best use it, from making the migration from X, to learning Bluesky's quirks, to finding the people you used to follow elsewhere. But as much as it can feel like Old Twitter, Bluesky has some unique things to offerchiefly among them a lack of an algorithm that "learns" (or decides) what you're interested in. That's where a tool in the independently created Bluesky Directory comes in handy.Search more than 47,000 Bluesky "starter packs"Assembled by Mubashar Mubs Iqbal, Bluesky Directory is essentially a collection of links to Bluesky tools and utilities from across the internet, from a Chrome extension that helps you find all of your old X followers to links to third-party clients that are more customizable than the official Bluesky app and website. The most useful part of the directory, for Bluesky newbies at least, will be the searchable index of more than 47,000 "starter packs." If you haven't encountered one yet, a starter pack is essentially a list of Bluesky accounts sorted by what unifies the folks behind them. When you open a starter packeither within the Bluesky app or on the websiteyou can choose to follow everyone in it with a single tap or click, or scroll down and choose who you want to appear in your feed one by one.So, for example, if you're a big movie geek, you might enjoy a starter pack like Filmmakers, which features 20-odd directors (Rian Johnson, Kevin Smith, and Lily Wachowski among them). If you love podcasts, you could check out Women in Podcasting. If you want to keep tabs on everyone who writes for Lifehacker, I've created a starter pack for that too.You can search for starter packs on Bluesky itself, but the experience isn't great. Bluesky Directory's searchable index is a lot more user-friendly. You just type a few relevant terms into the search bar, and you'll be served up any starter packs that match (the site is slightly sluggishwhich is to be expected with a million new users coming online every daybut not unusably so). You'll be able to see a description, the number of accounts on the list, and even a preview of a few of them, right from the results page. There's a starter pack for (almost) everythingYou can browse a list of the most popular starter packs, but if you want a truly bespoke social media experience, it pays to do a granular search. As of this writing, the index includes 47,450 different options, with more being added every minute. (If you've created a starter pack you'd like added to the database, there's a link to submit one right there on the search page). And chances are good if you're interested in a topic, there's a starter pack for it. Perhaps you want to follow authors who like to post about their cats? Or maybe academics with learned opinions about pizza? COVID-conscious porn performers? I was disappointed there were no results for a search for "Ninja Turtles," but it's early days yet; on the other hand, I found lists of journalists who write for some of my favorite publications, as well as people who've guested on my favorite movie podcast.Once you find a feed you like, click on the name and you'll be taken to Bluesky, where you can choose who on the list you want to follow (or follow everyone). You can also toggle over to the "Posts" tab to see recent posts from everyone in the starter pack. Before you know it, you'll have a hyper-specific list of Bluesky follows attuned to your particular interestswhich seems, to me, a lot better than having an algorithm throwing rage-bait at your face every time you open the app.

Two undersea communications cables in the Baltic Sea have been knocked offline, and at least one appears to have been physically cut. CNN received confirmation from a local telecom company that a cable between Lithuania and Sweden was cut on Sunday morning. A second cable, about 60 to 65 miles from the first, routes communications between Finland and Germany. The cause of that outage has yet to be determined, but officials suspect intentional damage.The outages follow a September warning from the US about an increased risk of Russian sabotage of undersea cables. That came after a joint investigation from public broadcasters from Sweden, Denmark, Norway and Finland that Russia had deployed a fleet of spy ships in Nordic waters. They were reportedly part of a program designed to sabotage the cables (and wind farms).This doesnt leave the European nations entirely without online communications, as data is typically routed through multiple cables to avoid overreliance on a single one.Cinia, the state-controlled Finnish company that oversees the second cable, said it wasnt yet determined what caused the outage since they havent yet physically inspected it. However, the sudden outage reportedly suggests it, too, was cut by an outside force.The foreign ministers of Finland and Germany released a joint statement on Monday. We are deeply concerned about the severed undersea cable connecting Finland and Germany in the Baltic Sea, they wrote. The fact that such an incident immediately raises suspicions of intentional damage speaks volumes about the volatility of our times. A thorough investigation is underway. Our European security is not only under threat from Russias war of aggression against Ukraine, but also from hybrid warfare by malicious actors. Safeguarding our shared critical infrastructure is vital to our security and the resilience of our societies.The Lithuania-Sweden cable, which handles about a third of Lithuanias internet capacity, is expected to be repaired over the next few weeks, and weather could determine the precise timing.This article originally appeared on Engadget at https://www.engadget.com/computing/two-baltic-sea-communications-cables-have-been-knocked-offline-214130723.html?src=rss

A four-pack of Samsung Galaxy SmartTag2 trackers are on sale via Amazon for just under $62. This Black Friday deal brings this particular combo pack down to a record low price. The typical cost is $100, so this represents a discount of 38 percent. This is one of our favorite Bluetooth trackers, and with good reason. They are the perfect companion accessory to Samsung phones. The finding network is vast and the battery life is fantastic. These trackers will stay powered on for up to 500 days, or even longer with Power Saving Mode enabled. We also appreciate the simple, yet aesthetically pleasing, design. Theres an actual hole for keychains, which is something Apple AirTags lack. These trackers are IP67-rated for water and dust-resistance. This means that the unit is protected from dust thats larger than 1mm in diameter and that it can be submerged in water up to one meter deep for up to 30 minutes. Just dont throw it in a lake and you should be good to go. The four-pack includes two white tags and two black tags, for a little yin and yang action. Theres not really a serious downside here for Samsung users. However, theres a major caveat for everyone else. The SmartTag2 trackers only work with Samsung phones and tablets. Check out all of the latest Black Friday and Cyber Monday deals here.This article originally appeared on Engadget at https://www.engadget.com/deals/a-four-pack-of-samsung-galaxy-smarttag2-trackers-is-38-percent-off-for-black-friday-194307496.html?src=rss

Looking for Quordle clues? We can help. Plus get the answers to Quordle today and past solutions.

Looking for NYT Strands answers and hints? Here's all you need to know to solve today's game, including the spangram.

Super Micro said on Monday that it hired BDO as its independent auditor.

Canva's new finance chief Kelly Steckelberg helped take Zoom public five years ago.

Its a great chance for us to challenge ourselves while giving back and doing something for others. Meet Sam Sharplin, a Models Supervisor at Wt FX and co-founder of the Movember team. Watch Sam share how our team is making a difference for mens mental health and beyond #WtFX #VFX #InternationalMensDay2024