投稿時間:2023-06-05 20:24:07 RSSフィード2023-06-05 20:00 分まとめ(30件)

カテゴリー等	サイト名等	記事タイトル・トレンドワード等	リンクURL	頻出ワード・要約等/検索ボリューム	登録日
IT	気になる、記になる…	Appleを装ったフィッシングメールの報告が増えているとしてフィッシング対策協議会が注意喚起	https://taisy0.com/2023/06/05/172523.html	apple	2023-06-05 10:38:07
IT	InfoQ	Windows Dev Drive - Storage Volume Customized for Developers with Improved Performance	https://www.infoq.com/news/2023/06/windows-dev-drive/?utm_campaign=infoq_content&utm_source=infoq&utm_medium=feed&utm_term=global	Windows Dev Drive Storage Volume Customized for Developers with Improved PerformanceMicrosoft released Dev Drive at the Build developer conference a custom storage volume geared for developers Built on top of Microsoft s proprietary ReFS Resilient File System Dev Drive is optimized for heavy I O operations and has improved performance and security capabilities By Giorgi Dalakishvili	2023-06-05 10:28:00
IT	ITmedia 総合記事一覧	[ITmedia News] 緑茶はグラスの“厚さ”で味が変わる？　中央大らが発表　飲み口が厚いと甘く、薄いと苦くなる	https://www.itmedia.co.jp/news/articles/2306/05/news164.html	itmedia	2023-06-05 19:16:00
python	Pythonタグが付けられた新着投稿 - Qiita	SCIP ソルバー + Python + PySCIPOpt パッケージ で、手っ取り早く数理最適化モデルを構築して結構速く解く、ということが仕事でもタダでできそう	https://qiita.com/samuelladoco/items/a629a24901f91f832cdb	apachelicense	2023-06-05 19:24:57
js	JavaScriptタグが付けられた新着投稿 - Qiita	【JavaScript】DataTablesが超便利すぎる	https://qiita.com/twrcd1227/items/d3f8db42235d70d24bf7	datatables	2023-06-05 19:01:18
Ruby	Rubyタグが付けられた新着投稿 - Qiita	オブジェクト指向で作りたい！クラス実装やリファクタリングする時のヒント	https://qiita.com/tatsumi_t2/items/df6f20f11fb89d70e773	開発	2023-06-05 19:25:20
Ruby	Rubyタグが付けられた新着投稿 - Qiita	Rubyのp、pp、puts、printメソッドの違いについて	https://qiita.com/Uts/items/1e744f506227d2cca809	print	2023-06-05 19:05:47
Docker	dockerタグが付けられた新着投稿 - Qiita	Mac のSystem ボリュームがどうしてもデカすぎて困っている人へ	https://qiita.com/kenmaro/items/09686815cad55aea33fd	docker	2023-06-05 19:24:14
Docker	dockerタグが付けられた新着投稿 - Qiita	ジェネリックDockerを実装する際のセキュリティ	https://qiita.com/kotobuki5991/items/87651ac15b848500ae4c	docker	2023-06-05 19:11:30
golang	Goタグが付けられた新着投稿 - Qiita	ジェネリックDockerを実装する際のセキュリティ	https://qiita.com/kotobuki5991/items/87651ac15b848500ae4c	docker	2023-06-05 19:11:30
技術ブログ	Developers.IO	Amazon RDS for MySQL でいつからレプリケーションエラーが発生しているかを確認したい	https://dev.classmethod.jp/articles/find-out-when-the-replication-error-occurred/	amazonrdsformysql	2023-06-05 10:23:10
技術ブログ	Developers.IO	韓国で今話題のスターバックス店舗でサイレンオーダーをやってみた	https://dev.classmethod.jp/articles/trying-siren-order-in-korea/	韓国	2023-06-05 10:22:48
技術ブログ	Developers.IO	AWS Organizations で使用する AWS CloudTrail の Amazon CloudWatch Logs への出力を委任管理アカウントのコンソールから有効化できない	https://dev.classmethod.jp/articles/cloudtrail-design-account-enabled-logs-for-console/	cloud	2023-06-05 10:10:31
技術ブログ	Developers.IO	AWS CDKでAmazon Aurora DB クラスターとAWS Secrets Managerとの統合を設定してみた	https://dev.classmethod.jp/articles/aws-cdk-aws-rds-integration-aws-secrets-manager/	amazonsecretsmanager	2023-06-05 10:06:03
海外TECH	MakeUseOf	8 Hardware Features That Will Make iPhones Even Better	https://www.makeuseof.com/hardware-features-that-will-make-iphone-better/	android	2023-06-05 10:15:17
海外TECH	DEV Community	Howto turn a x86 binary executable back into C source code	https://dev.to/torbenfeldthusen/howto-turn-a-x86-binary-executable-back-into-c-source-code-45nd	Howto turn a x binary executable back into C source codeObjective turn a x binary executable back into C source code Understand how the compiler turns C into assembly code Low level OS structures and executable file format Arithmetic Instructionsmov eax eax mov ebx ebx add eax ebx eax eax ebx sub ebx ebx ebx Accessing Memorymox eax eax int mov ebx ebx mov eax ebx eax ebx mov ebx eax ebx eax Conditional Branchescmp eax compare eax with je label if eax goto label ja label if eax gt goto labeljb label if eax lt goto label jbe label if eax lt goto label jne label if eax goto label jmp label unconditional goto labelFunction callsFirst calling a function call func store return address on the stack and jump to func The first operations is to save the return pointer pop esi save esi Right before leaving the function pop esi restore esiret read return address from the stack and jump to it Modern Compiler ArchitectureC code gt Parsing gt Intermediate representation gt optimization gt Low level intermediate representation gt register allocation gt x assemblyHigh level OptimizationsInliningFor example the function c int foo int a int b return a b c foo a b translates toc a b Loop unrollingThe loop for i i lt i a i becomes a a Loop invariant code motionThe loop for i i lt i a i p q becomes temp p q for i i lt i a i temp Common subexpression eliminationThe variable attributions Objective turn a x binary executable back into C source code Understand how the compiler turns C into assembly code Low level OS structures and executable file format Arithmetic Instructionsmov eax eax mov ebx ebx add eax ebx eax eax ebx sub ebx ebx ebx Accessing Memorymox eax eax int mov ebx ebx mov eax ebx eax ebx mov ebx eax ebx eax Conditional Branchescmp eax compare eax with je label if eax goto label ja label if eax gt goto labeljb label if eax lt goto label jbe label if eax lt goto label jne label if eax goto label jmp label unconditional goto labelFunction callsFirst calling a function call func store return address on the stack and jump to func The first operations is to save the return pointer pop esi save esi Right before leaving the function pop esi restore esiret read return address from the stack and jump to it Modern Compiler ArchitectureC code gt Parsing gt Intermediate representation gt optimization gt Low level intermediate representation gt register allocation gt x assemblyHigh level OptimizationsInliningFor example the function c int foo int a int b return a b c foo a b translates toc a b Loop unrollingThe loop for i i lt i a i becomes a a Loop invariant code motionThe loop for i i lt i a i p q becomes temp p q for i i lt i a i temp Common subexpression eliminationThe variable attributions a b z p q z becomestemp z a b zp q zConstant folding and propagationThe assignments a b a func b Becomes func Dead code eliminationDelete unnecessary code a if a lt printf “ERROR toa Low Level OptimizationsStrength reductionCodes such as y x y x Becomes y x xy x lt lt xCode block reorderingCodes such as if a lt goto lprintf “ERROR goto labell printf “OK l return Becomes if a gt goto lprintf “OK l returnl printf “ERROR goto lRegister allocationMemory access is slower than registers Try to fit as many as local variables as possible in registers The mapping of local variables to stack location and registers is not constant Instruction schedulingAssembly code like mov eax esi add eax mov ebx edi add ebx Becomes mov eax esi mov ebx edi add eax add ebx a b z p q z becomestemp z a b zp q zConstant folding and propagationThe assignments a b a func b Becomes func Dead code eliminationDelete unnecessary code a if a lt printf “ERROR toa Low Level OptimizationsStrength reductionCodes such as y x y x Becomes y x xy x lt lt xCode block reorderingCodes such as if a lt goto lprintf “ERROR goto labell printf “OK l return Becomes if a gt goto lprintf “OK l returnl printf “ERROR goto lRegister allocationMemory access is slower than registers Try to fit as many as local variables as possible in registers The mapping of local variables to stack location and registers is not constant Objective turn a x binary executable back into C source code Understand how the compiler turns C into assembly code Low level OS structures and executable file format Arithmetic Instructionsmov eax eax mov ebx ebx add eax ebx eax eax ebx sub ebx ebx ebx Accessing Memorymox eax eax int mov ebx ebx mov eax ebx eax ebx mov ebx eax ebx eax Conditional Branchescmp eax compare eax with je label if eax goto label ja label if eax gt goto labeljb label if eax lt goto label jbe label if eax lt goto label jne label if eax goto label jmp label unconditional goto labelFunction callsFirst calling a function call func store return address on the stack and jump to func The first operations is to save the return pointer pop esi save esi Right before leaving the function pop esi restore esiret read return address from the stack and jump to it Modern Compiler ArchitectureC code gt Parsing gt Intermediate representation gt optimization gt Low level intermediate representation gt register allocation gt x assemblyHigh level OptimizationsInliningFor example the function c int foo int a int b return a b c foo a b translates toc a b Loop unrollingThe loop for i i lt i a i becomes a a Loop invariant code motionThe loop for i i lt i a i p q becomes temp p q for i i lt i a i temp Common subexpression eliminationThe variable attributions Objective turn a x binary executable back into C source code Understand how the compiler turns C into assembly code Low level OS structures and executable file format Arithmetic Instructionsmov eax eax mov ebx ebx add eax ebx eax eax ebx sub ebx ebx ebx Accessing Memorymox eax eax int mov ebx ebx mov eax ebx eax ebx mov ebx eax ebx eax Conditional Branchescmp eax compare eax with je label if eax goto label ja label if eax gt goto labeljb label if eax lt goto label jbe label if eax lt goto label jne label if eax goto label jmp label unconditional goto labelFunction callsFirst calling a function call func store return address on the stack and jump to func The first operations is to save the return pointer pop esi save esi Right before leaving the function pop esi restore esiret read return address from the stack and jump to it Modern Compiler ArchitectureC code gt Parsing gt Intermediate representation gt optimization gt Low level intermediate representation gt register allocation gt x assemblyHigh level OptimizationsInliningFor example the function c int foo int a int b return a b c foo a b translates toc a b Loop unrollingThe loop for i i lt i a i becomes a a Loop invariant code motionThe loop for i i lt i a i p q becomes temp p q for i i lt i a i temp Common subexpression eliminationThe variable attributions a b z p q z becomestemp z a b zp q zConstant folding and propagationThe assignments a b a func b Becomes func Dead code eliminationDelete unnecessary code a if a lt printf “ERROR toa Low Level OptimizationsStrength reductionCodes such as y x y x Becomes y x xy x lt lt xCode block reorderingCodes such as if a lt goto lprintf “ERROR goto labell printf “OK l return Becomes if a gt goto lprintf “OK l returnl printf “ERROR goto lRegister allocationMemory access is slower than registers Try to fit as many as local variables as possible in registers The mapping of local variables to stack location and registers is not constant Instruction schedulingAssembly code like mov eax esi add eax mov ebx edi add ebx Becomes mov eax esi mov ebx edi add eax add ebx a b z p q z becomestemp z a b zp q zConstant folding and propagationThe assignments a b a func b Becomes func Dead code eliminationDelete unnecessary code a if a lt printf “ERROR toa Low Level OptimizationsStrength reductionCodes such as y x y x Becomes y x xy x lt lt xCode block reorderingCodes such as if a lt goto lprintf “ERROR goto labell printf “OK l return Becomes if a gt goto lprintf “OK l returnl printf “ERROR goto lRegister allocationMemory access is slower than registers Try to fit as many as local variables as possible in registers The mapping of local variables to stack location and registers is not constant Instruction schedulingAssembly code like mov eax esi add eax mov ebx edi add ebx Becomes mov eax esi mov ebx edi add eax add ebx Instruction schedulingAssembly code like mov eax esi add eax mov ebx edi add ebx Becomes mov eax esi mov ebx edi add eax add ebx	2023-06-05 10:36:47
海外TECH	DEV Community	Astro Coding Session: "Series" feature on my blog	https://dev.to/this-is-learning/astro-coding-session-series-feature-on-my-blog-4e0p	Astro Coding Session quot Series quot feature on my blogHello If you read my articles often here on devto you probably noticed that I also publish them on my website leonardomontini dev It s a static Astro website hosted on Netlify I m using Typescript and the articles are stored in mdx files so I can easily add components inside them for example the YouTube video player The repository is Open Source you can have a look on GitHub Balastrong blog astro Source code of my homepage Blog HomeHello This is the source code of my homepage built with Astro You can find the deployed version at View on GitHubIf you re curious about how I code I recorded a short coding session where I added an initial version of the Series feature we have here on devto Basically I add a series field in the frontmatter of the mdx files and then I show in the article s page all the other articles of the same series Definitely not rocket science but I think it can be interesting to see how other developers actually work and how they solve problems What you will see in this session is GitHub Copilot doing of the work and myself struggling with tailwind classes Spoiler This week s article is quite short and the video is a simple coding session I hope you like it anyway But the spoiler is I m also working on a much longer article about custom GitHub Actions which I ll try to release next week The article is already x longer than usual I hope it won t take ages to record the video session Stay tuned Thanks for reading this article I hope you found it interesting I recently launched my Discord server to talk about Open Source and Web Development feel free to join Do you like my content You might consider subscribing to my YouTube channel It means a lot to me ️You can find it here Feel free to follow me to get notified when new articles are out Leonardo MontiniFollow I talk about Open Source GitHub and Web Development I also run a YouTube channel called DevLeonardo see you there	2023-06-05 10:35:08
海外TECH	DEV Community	Top GitHub repositories to learn modern React development	https://dev.to/livecycle/top-github-repositories-to-learn-modern-react-development-5d3h	Top GitHub repositories to learn modern React development TL DROpen source is great for many things One of them is learning new skills In this article we ll look at some of the best Open Source React projects on GitHub that you can use to quickly boost your hands on React learning and coding experience React tutorials will only take you so farThere are plenty of React tutorials available online Many of them are excellent resources for teaching you the fundamentals of React But where most courses fall short is when it comes to explaining how the different pieces of a modern React app tie together When first working on a large application many developers find themselves struggling with questions such as “How do I efficiently manage my folder structure or “How can I pass state across a big application These are the types of “real life coding issues that have been better tackled outside the classroom by teams that have actually worked together on building larger apps Open Source A Free UniversityOne way to develop these types of skills is to join such a team and see for yourself how the senior developers work on the application But as simple as this approach sounds these opportunities can seem few and far between especially if you are at the beginning of your journey and you haven t yet secured a job Thankfully there s a free alternative that doesn t require experience or an interview Open Source Open Source repositories enable you to get involved in large scale projects immediately It s almost like a back door cheat code to join these teams as a developer and get feedback from other devs by contributing to the project One of the many reasons we love Open SourceThis is one of the many reasons we love open source Not only is it great for the projects and the maintainers themselves by getting contributions and input from the community but it s also a tremendous reciprocal benefit to all community members at large Open source gives every developer immediate access to information feedback and hands on experience that can help them grow and progress These are just a few of the reasons we ve been so excited about out the launch of our open source tool “Preevy Not only is it something that development teams can benefit from by easily creating shareable preview environments but it also has been a great way for us to engage with other developers in the community who are interested to learn from and contribute to a project such as this Feel free to check it out leave us a star and make a contribution️️️ So with that context let s settle in and look at some of the best Open Source React projects on GitHub that you can use to quickly boost your hands on React learning and experience Open Source React projects you can learn from Cal comCal com labels itself as “Scheduling infrastructure for absolutely everyone They re a competitor to appointment scheduling services like Calendly They have both a hosted and self hosted offering It s a full stack Next js app that relies on tRPC for typesafe client to server communication The repository has quite an extensive Monorepo setup using Turborepo to split the app up into various packages On the styling front Cal com uses TailwindCSS and headless Radix components The repository is very active and they actively encourage contributors Many issues are labeled as “good first issue and “help wanted TaxonomyThis one is a little bit different It was built by a single developer as a demo to explore the new Next js App Router features But don t let that fool you It s arguably one of the most comprehensive Open Source apps out there that use the new App Router So this is a great resource even for experienced React developers who want to learn how to use the App Router in production Besides Next js for both front end and backend Taxonomy uses NextAuth for authentication Prisma as the ORM and TailwindCSS with Radix for styling Particularly interesting is the Stripe subscription integration So if you d like to learn how to integrate subscriptions into your app this is the repository for you HighstormHighstorm is a new player on the block It s a service to monitor events that happen in your application You submit events through their API which then get fed into the Highstorm dashboard Once again this is a project based on Next js It uses Tinybird as the database for the analytics data and Clerk for authentication This project is great if you want to learn how to process large amounts of analytics data and display them in fancy charts Dub shDub sh was launched roughly a year ago as an Open Source link shortener It s an alternative to services like Bitly It is also based on Next js and uses a combination of the old Pages and the new App router It s usually one of the first projects to adopt new Next js features such as the Metadata API The repository is a great place to learn about multi tenancy Next apps These are apps that serve different users under different domains In the case of Dub sh users can add their own domain to create shortened links under The app is also beautifully designed with many joyful animations all over the site The Framer Motion library is used to help with this Bulletproof ReactBulletproof React is different compared to the other projects we have covered so far The repository does contain a fully functioning demo React app But the most value comes from reading the comprehensive documentation that comes with this app This documentation lays out best practices for all aspects of designing a large scale React app This is a great resource to reference on a regular basis to refresh your memory on the best practices Wrapping up and getting startedAs we have seen each of the above repositories provides a unique perspective into how large scale React applications are built Everyone can benefit from studying the code of the repositories no matter how experienced you are A great practical tip for starting to learn from a repository is to pose a specific question For example you might ask something like “how does Taxonomy handle Stripe subscriptions Then dig through the layers of code to find an answer to your question And one last point exploring Open Source repositories is a great way to level up your React skills but it s not the only thing you should do It is important to also apply your newly gained knowledge to your own projects That s how you ll solidify new skills and retain the information The main thing is to enjoy the process and recognize that open source is a friendly place to develop new skills no matter what your current level of experience is	2023-06-05 10:31:59
海外TECH	DEV Community	Fear and the Future: AI's Threat to job security - An interview with AI professor Alejandro Piad Morffis	https://dev.to/edemgold/fear-and-the-future-ais-threat-to-job-security-an-interview-with-ai-professor-alejandro-piad-morffis-meb	Fear and the Future AI x s Threat to job security An interview with AI professor Alejandro Piad Morffis“It would be pretty damn maddening if it turns out programmers are easier to automate than lawyers Professor Alejandro Piad MorffisThe increase in adoption of Large Language Generative AI models such as ChatGPT Microsoft Bing Google Bard Stable Diffusion etc while the advantages of these models cannot be refuted it has led to an exaggerated and harrowing but not baseless fear by members of the public on the possibility of these AI models jeopardizing job security for millions of workers worldwide As described earlier the threat of AI to human jobs while being exaggerated and harrowing isn t baseless The ability of AI to perform repetitive tasks and process large amounts of information and mimic human like decision making makes it a very good tool to enhance creativity productivity and efficiency To answer the question will AI take our jobs I have enlisted the help of an expert by the name of Professor Alejandro Piad Morffis A Professor of AI at the University of Havana Cuba The Professor is a mentor teacher friend and most importantly inspiration to me How I hope to approach thisThe questions will be pre fixed with the Q letter while the answers will be pre fixed with the A letter With regards to the questions I hope to cover technical and philosophical questions as Professor Morffis also has an affinity for the Philosophical It is also important to note that I will provide links to certain concepts that are complex to grasp for the sake of understanding Let us Begin Q Firstly Could you tell us a bit about yourself your professional qualifications and suchA My name is Alejandro Piad I majored in Computer Science at the School of Math and Computer Science at the University of Havana Cuba I did a Master s in Computer Science also at the same college in and earned a double PhD in Computer Science at the University of Alicante and a PhD in Math at the University of Havana in My PhD was in knowledge discovery from natural language specifically focused on entity and relation extraction from medical text Since grad school I ve been teaching at the University of Havana I ve been the main lecturer in Programming Compilers and Algorithm Design and also an occasional lecturer on Machine Learning and other subjects Since I m a full time professor there I was also one of the founders of the new Data Science career the first of its kind in Cuba and I wrote the entire Programming and Computing Systems curriculum for that career I keep doing research in NLP right now focusing on neuro symbolic approaches to knowledge discovery mixing LLMs with symbolic systems Q How long have you been working with AI systems A I played with AI for games as an undergrad student and did a couple of student projects with computer vision and metaheuristics After graduating I started my master s in Computer Graphics but as a side project I did some research in NLP specifically on sentiment analysis on Twitter After finishing the master I started thinking about doing a PhD and got all in with machine learning So you could say around years since I started taking AI seriously My oldest paper related to this stuff is around Q That is intensely impressive You worked with AI way before it became cool What do you believe is the singular most significant technical advancement in AI which has contributed to its current mainstream adoption and the consequent job displacement threats A Well it was always cool just not outside academia I d say the intersection of two orthogonal developments the discovery of artificial neural network architectures such as the Transformer which solved many of the scalability problems of previous architectures and the invention of hardware where you can run those specific architectures at scale super efficiently Q Fascinating In your professional opinion as an educator and an AI researcher what industries stand the risk of being replaced by AI A I don t know if any industry will be replaced entirely but I m sure there will be massive changes In the long term of course no one can say anything But in the short and mid term years with what we re seeing with language models my bet is that anyone whose job is predicated on the shallow filtering and processing of natural language will have some reckoning to do This includes all sorts of managerial roles including anyone whose job is to read emails summarize and build reports Any kind of secretary who doesn t go beyond note taking and task scheduling Copywriters who work with templated content Basically any content creation task below the level of actual human creativity will be cheaper to automate than paying a human stochastic parrot So those will go away One single copywriter using the ChatGPT of the near future will be able to craft in hypothesis x to x more content with the same quality Not because the model will give them the final quality they aim for but because the model will give them of the quality and then the real human creativity comes as a cherry on top and adds the final Education has to change considerably too We can talk more about that if you want Q This is a really nice angle You re an educationist and from your piece titled Rethinking college education you obviously know how change averse the educational institution is Do you think formal education can be depended upon for survival in the post AI world A Yeah academia will adapt It is the longest living institution in Western civilization It predates all our mainstream religions and it has survived all major civilization changes It will change substantially as it has changed across the ages Q How important is it for society to consider the ethical implications of AI and job displacement A All technology has potential issues and the more advanced the tech the more pressing it is to consider them AI is a very powerful technology with the potential to disrupt all our economic relationships It is something at the level of an industrial revolution so it will have massive implications and so the concern must be at the same level One thing that is different from previous disruptive tech is that mostly new tech automates the jobs that require the least cognitive skills it happened with agriculture manufacturing mining etc However this time we are on the brink of replacing a large number of white collar jobs while leaving lots of blue collar jobs undisrupted so we will have lots of people who are used to working in offices finding that an AI can do their job as well or maybe slightly better and much cheaper so they will either have to upgrade their skills significantly or they will have to turn to less skilled jobs there are other ethical considerations there is a lot of potential for misuse of AI technologies for misinformation fake news social disruption etc I don t think we are prepared for a massive number of human like chatbots taking over Twitter it is already starting to happen There are also bias issues as these systems become more and more pervasive the harms can be very focused on the minorities so everyone will not reap the benefits of AI to the same degree some minorities will get the downsides more strongly than those not from minorities Q So in other words we should pay attention because unlike past forms of automation AI has the potential to disrupt cognitively tasking jobs as well A Yeah especially those jobs It will automate more white collar jobs than blue collar jobs at least in the near term That s something new and society isn t used to having to deal with that kind of job disruption These are folks that went to college and more or less got convinced their jobs were safe or at least safer than taxi drivers pizza boys gardeners you mention it Q This makes sense let s hit a little bit close to home Do you believe that an increase in AI capability will ultimately lead to a decrease in overall employment for software engineers developers A In the very long term all jobs will evolve in unpredictable ways including software engineering and development AI and technological advancements will transform these professions to the point where they may seem to have disappeared However in the short to midterm a decrease in software engineers is unlikely due to the increasing demand for software across various industries This growing need for skilled professionals far surpasses the current number of trained individuals capable of building software The AI revolution will follow a similar pattern as previous technological breakthroughs in computer science such as compilers integrated development environments cloud computing containers code completion and IntelliSense These innovations made programming more accessible for those without highly formal backgrounds and expanded opportunities for developers Over the next years we can expect an explosion of people entering the field of software development Although job roles may change somewhat with evolving technology trends there will likely be continued growth for those interested in learning how to program and write code Q This is incredible although the release of Generative AI models such as GitHub s Co Pilot and the GPT family of models has prompted forgive my pun rumours about the possibility of software developers losing their jobs to AI What do you say about this A Look at the numbers All I m seeing are more job ads for software developers The trend is still climbing Q Jeff Clune an ex OpenAI engineer recently made a prediction at the AI Safety Debate conference he stated that there was a percent chance that AI will be capable of handling of economically valuable work by the year what would this mean for the overall developer labour market A First I have no idea how you would wrap your head around what a chance of automating a of jobs even looks like Is it a expectation of losing your job Q I guess the numbers do make for a confusing scene But the essential point is Software developers have lots of reasons to be worried about their job security and many of the tasks they currently spend lots of time on are being automated The pace at which that occurs will accelerate A Yeah but the thing is many of the tasks developers spend most of their time on are pretty low value and we would be much better off if they were automated debugging writing tests doing pesky code optimizations As you automate all of that we ll have more time for the really important parts of software development which was never about writing code Q Could you speak more about those parts A High level and architecture design user experience human computer interaction and that s just about the software itself Software engineering is really about the relationship between software and people both people that make software and people that use software So software skills are only half of the story Understanding your users and colleagues is the other half Q This is similar to how the core job of accountants is to communicate financial information and not create financial statements fascinating It s a fair bet to say AI capabilities will increase in years How prepared are we as a society amp species to address the potential job displacement loss brought on by the potential adoption of AI How does this affect our sense of purpose as human beings A Very hard to say of course we re in the middle of an industrial revolution as big as at least the microprocessor revolution or the internet revolution no one in could imagine what would look like Society is never ready for change by definition That s what a system is something that strives to maintain its status quo But humans are the most adaptable social species out there so I think we ll manage Lots of people will suffer and that s something we have to work on definitely but nothing apocalyptic in my opinion will happen Q There have been a lot of talks on the dystopian potential of AI Why do you say nothing apocalyptic will occur A I still haven t seen any really compelling arguments for the doomsday scenario Lots of the arguments seem to be predicated on reasoning like we don t know how this is going to evolve so it will probably kill us all and that s a classic logical fallacy you re basically making an inference from lack of knowledge Q This is true But the AI alignment problem does seem plausible A I think we will solve it at least well enough to avoid apocalyptic scenarios The most severe alignment issues require you to believe in a powerful version of the orthogonality thesis that I don t believe plausible SubscribedQ Fascinating Going back to automation How can we leverage AI to augment human work rather than replace it and what industries are ripe for this kind of collaboration A I think that s only natural as we automate more and more of the menial cognitive work e g summarizing documents or finding relevant references we humans will get to work on the most creative parts of our jobs Some jobs have very little of that to begin with and there I see a challenge because maybe those will be completely or almost completely automated away But most knowledge work has a creative side the part where you actually do something novel As to which fields are ripe for this I can t talk about much else but in education at least I think we re bound for a long needed revolution We professors no longer need to be gatekeepers of information Instead of spending most of our time grading the same essays over and over we can now focus on giving the best possible personal feedback to each student Q What are the possible ways AI could revolutionise the educational system Perhaps more teaching techniques adapted optimally to students A There are a few easy ways and then some not that easy First is just a matter of increasing access to knowledge Now almost everything you want to learn you can find relevant information on the internet at least to begin with but it is often split around many sources with disparate levels of detail contradictory stuff different linguistic styles etc The first relatively easy application is just here take this bunch of sources on some topic and give me a high level overview of the main takeaways summarized with links to dive deeper etc We are pretty close to that baring the hallucinations which are a significant problem Another way is by simply freeing educators from menial tasks to give them more time to focus on creating learning experiences But by far the most important thing I believe is the potential for personalized learning You could have an AI assistant and tell it I want to learn how to make a rocket and it could create a very detailed plan especially for you based on what it already knows that you know it would tell you here first watch this video now take this short course now read this chapter of this book And guide you for months to learn something very specific Q This is truly promising You make a solid case for humans adapting You spoke about bias is it fair to say large scale AI adoption will affect the minority If yes how can this be combatedA Yeah definitely machine learning is by definition trained on the majority so it will always hurt the most those whose use case doesn t fit the majority for any reason In particular whenever you train models to predict human behaviour or interact with humans it tends to work better for the subpopulations that are best represented in the data What can you do Start by raising awareness of these issues and make sure to thoroughly test your models for bias Be very careful about how you collect data don t go the easy way and crawl the web and make an effort to find good high quality and high diversity sources of data But more than anything include diverse people with diverse points of view in your team You can t solve a problem you can t see Q This makes me think Is there a possibility that access to said AI tools will be relegated to the financially capable A I m hoping the open source community will make the tools available to all We already have seen what having access to a free operating system a free office suite a free game engine a free code editor etc does for the creative kids of the poorer parts of the world I trust we will have open source AI tools as good as commercial ones the same way we have open source dev tools as good as commercial ones Q This seems really feasible What advice would you give to your students to prepare them for the workforce in a post AI world A If you are already studying computer science the basic advice is to focus on fundamentals not just tools Tools will change but the fundamentals will remain relevant for a long time If studying something else learn how AI can improve your productivity and learn a lot about its limitations Use it to make your own work better Q This makes sense fundamentals will stand the test of time Thank you very much for your time Professor Morffis Any closing words A The AI revolution is here We can all either be a part of it by learning to use this technology for making good and improving the lives of everyone If you enjoyed this article Invest in the writer	2023-06-05 10:01:06
ニュース	BBC News - Home	Ukraine war: Russia says it thwarted major Ukrainian offensive	https://www.bbc.co.uk/news/world-europe-65806152?at_medium=RSS&at_campaign=KARANGA	ukrainian	2023-06-05 10:53:39
ニュース	BBC News - Home	'Instagram seller quoted me £500 for a GCSE paper'	https://www.bbc.co.uk/news/education-65768038?at_medium=RSS&at_campaign=KARANGA	boards	2023-06-05 10:18:38
ニュース	BBC News - Home	Bournemouth beach deaths: Pair 'caught in riptide' - inquest	https://www.bbc.co.uk/news/uk-england-dorset-65807339?at_medium=RSS&at_campaign=KARANGA	bournemouth	2023-06-05 10:52:56
ニュース	BBC News - Home	France: Paris Champs-Élysées hosts mass spelling contest	https://www.bbc.co.uk/news/world-europe-65809523?at_medium=RSS&at_campaign=KARANGA	dictation	2023-06-05 10:09:28
ニュース	BBC News - Home	Appeal over Vivienne Westwood planter urn theft from grave	https://www.bbc.co.uk/news/uk-england-derbyshire-65810416?at_medium=RSS&at_campaign=KARANGA	tintwistle	2023-06-05 10:31:03
ニュース	BBC News - Home	French Open 2023: Ons Jabeur into quarter-finals for first time with win over Bernarda Pera	https://www.bbc.co.uk/sport/tennis/65810014?at_medium=RSS&at_campaign=KARANGA	French Open Ons Jabeur into quarter finals for first time with win over Bernarda PeraTunisia s Ons Jabeur reaches the quarter finals of the French Open for the first time with a win over American Bernarda Pera	2023-06-05 10:31:53
ビジネス	不景気.com	クックパッドが退職勧奨・解雇で110名を削減へ、今年3度目 - 不景気com	https://www.fukeiki.com/2023/06/cookpad-cut-110-job.html	退職勧奨	2023-06-05 10:20:44
IT	週刊アスキー	『ウマ娘』ウイニングライブはプロの情熱の結晶！アクター＆アーティスト座談会記事が公開	https://weekly.ascii.jp/elem/000/004/139/4139674/	cygames	2023-06-05 19:10:00
IT	週刊アスキー	「メルカリShops」、名古屋市など8つの自治体がまだ使える備品や粗大ごみを販売開始	https://weekly.ascii.jp/elem/000/004/139/4139634/	shops	2023-06-05 19:15:00
IT	週刊アスキー	バーチャル高校野球の無料ライブ配信、2023年は1回戦から全試合配信が決定！	https://weekly.ascii.jp/elem/000/004/139/4139673/	sportsbull	2023-06-05 19:45:00
IT	週刊アスキー	BASE、Pay IDであと払いを利用で国際緑化推進センターに寄付をする取り組みをスタート	https://weekly.ascii.jp/elem/000/004/139/4139681/	payid	2023-06-05 19:30:00