Artificial intelligence is rapidly changing many aspects of how we work and live. (How many stories did you read last week about self-driving cars and job-stealing robots? Perhaps your holiday shopping involved some AI algorithms, as well.) But despite the constant flow of news, many misconceptions about AI remain.
AI doesn’t think in our sense of the word at all, Scriffignano explains. “In many ways, it’s not really intelligence. It’s regressive.”
IT leaders should make deliberate choices about what AI can and can’t do on its own. “You have to pay attention to giving AI autonomy intentionally and not by accident,”
6,742 Comments
Tomi Engdahl says:
We can’t make this stuff up: Palantir, Anduril form fellowship for AI adventures
Meanwhile, Sauron secures funding. Really
https://www.theregister.com/2024/12/09/palantir_anduril_alliance/
Tomi Engdahl says:
OpenAI, why the new o1 models are the end of the GPT era
AI-powered marketing tools
OpenAI’s new o1 models autonomously evaluate and review their outputs before providing answers. Unprecedented computing power.
https://en.futuroprossimo.it/2024/12/openai-perche-i-nuovi-modelli-o1-sono-la-fine-dellera-gpt/
Tomi Engdahl says:
Maksatko ChatGPT:stä tai muista tekoälypalveluista?
08.12.2024 21:02 Muropaketin toimitus
ChatGPT:n voi valjastaa paremmin käyttöönsä kuukausimaksun avulla. Maksatko sinä ChatGPT:stä tai jostain muusta tekoälypalvelusta?
ChatGPT:n kohdalla kielimallin edistyneemmän Plus-tilauksen saa käyttöönsä 20 dollarilla kuussa. Lisäksi palvelusta on saatavilla vieläkin kattavampi uusi 200 dollaria kuussa maksava Pro-versio.
https://muropaketti.com/mobiili/mobiiliuutiset/gallup-maksatko-chatgptsta-tai-muista-tekoalypalveluista/
Tomi Engdahl says:
Getting tinyML Ready for Prime Time
Wake Vision, a new dataset for person detection, focuses on quality to squeeze the most performance possible out of tinyML models.
https://www.hackster.io/news/getting-tinyml-ready-for-prime-time-c95dfff4283f
Tomi Engdahl says:
Microsoft AI chief Mustafa Suleyman says conversational AI is the next web browser
The company’s new AI chief on working for Microsoft, the OpenAI relationship, and when superintelligence might actually arrive.
https://www.theverge.com/24314821/microsoft-ai-ceo-mustafa-suleyman-google-deepmind-openai-inflection-agi-decoder-podcast
Tomi Engdahl says:
DeepMind AI weather forecaster beats world-class system
Artificial-intelligence model provides forecasts 15 days out, as well as the probability of accuracy. And it does so faster than the best operational model.
https://www.nature.com/articles/d41586-024-03957-3
Tomi Engdahl says:
I can now run a GPT-4 class model on my laptop
9th December 2024
Meta’s new Llama 3.3 70B is a genuinely GPT-4 class Large Language Model that runs on my laptop.
Just 20 months ago I was amazed to see something that felt GPT-3 class run on that same machine. The quality of models that are accessible on consumer hardware has improved dramatically in the past two years.
My laptop is a 64GB MacBook Pro M2, which I got in January 2023—two months after the initial release of ChatGPT. All of my experiments running LLMs on a laptop have used this same machine.
https://simonwillison.net/2024/Dec/9/llama-33-70b/
Tomi Engdahl says:
More-powerful AI is coming. Academia and industry must oversee it — together
AI companies want to give machines human-level intelligence, or AGI. The safest and best results will come when academic and industry scientists collaborate to guide its development.
https://www.nature.com/articles/d41586-024-03911-3
Tomi Engdahl says:
https://www.fastcompany.com/91230209/chatgpt-chrome-browser-extension
Tomi Engdahl says:
Individual Improvements
Research has shown that individuals can increase their capabilities in a relatively short period of time. Studies of specialists such as customer-service agents, software engineers and data scientists found that gen AI significantly improved productivity with minimal training.
The degree of improvement can vary widely according to skill level, experience, and profession. In these studies, customer service agents resolved issues up to 34% faster, with new hires showing the most rapid improvement. Software engineers delivered 26% more code, and data scientists completed tasks, on average, in 10% less time.
While these are significant gains, they need to be put in perspective: Even a 34% one-time improvement for isolated tasks represents a much smaller impact when applied to an entire enterprise. When applied this way, gen AI represents what Nobel laureate Daron Acemoglu and his co-author Pascual Restrepo call “so-so technologies” — innovations that displace workers, but do not increase productivity enough to impact competitiveness or improve lives.
Many organizations are at this stage. It requires little more than making gen AI technology and guidance available to employees. At this stage, while there may be pockets of improvement, the effect is likely to be minimal. Depending on adoption rates in your industry, you might even find yourself lagging behind your peers.
https://hbr.org/2024/12/how-to-create-value-systematically-with-gen-ai
Tomi Engdahl says:
How AI agents will transform the future of work
https://www.infoworld.com/article/3611465/how-ai-agents-will-transform-the-future-of-work.html
AI agents are already reengineering software development, business processes, and customer experiences. Here’s what you need to know.
At first, robotic process automation coupled with low-code platforms and orchestration tools propelled many organizations to increase productivity and scale business operations. Virtual agents and chatbots took automation one step further by enabling a conversational experience. Then, large language models (LLMs), vector databases, retrieval augmented generation (RAG), and other generative AI innovations enabled new ways to summarize content, generate code using copilots, and answer questions conversationally.
AI agents combine automation, conversational experiences, and process orchestration capabilities to lead us to the next phase of generative AI evolution and digital transformation. They provide developers, business users, and others with a role-based partner, proactively automating steps and acting as knowledgeable collaborators in getting work done. Integrating genAI technologies with role-based workflows is a key opportunity to deliver transformational generative AI business benefits beyond productivity improvements.
Tomi Engdahl says:
“AI agents are changing the game across industries by automating tasks, solving problems, and improving workflows,” says Abhi Maheshwari, CEO of Aisera. “Unlike standard chatbots, these agents can reason, plan, and take action independently. They’re used in areas like tech, manufacturing, legal, retail, education, and government.”
Many platforms now have sidebars on webpages and other user experience elements where end-users can interact with AI agents around their work. Sometimes, the agent presents information proactively so people can take action. At other times, they lend expertise and share data-driven insights with the employee while performing their work.
https://www.infoworld.com/article/3611465/how-ai-agents-will-transform-the-future-of-work.html
Tomi Engdahl says:
Bing AI tells also references in text
Tomi Engdahl says:
How to generate unit tests with GitHub Copilot: Tips and examples
Learn how to generate unit tests with GitHub Copilot and get specific examples, a tutorial, and best practices.
https://github.blog/ai-and-ml/how-to-generate-unit-tests-with-github-copilot-tips-and-examples/
Developers writing enough unit tests? Sure, and my code never has bugs on a Friday afternoon.
Whether you’re an early-career developer or a seasoned professional, writing tests—or writing enough tests—is a challenge. That’s especially true with unit tests, which help developers catch bugs early, validate code, aid with refactoring, improve code quality, and play a core role in Test-Driven Development (TDD).
All of this to say, you can save a lot of time (and write better, more robust code) by automating your test generation—and AI coding tools are making that easier and quicker than ever.
GitHub Copilot, GitHub’s AI-powered coding assistant, helps generate test cases on the fly and can save you time. I’ll be honest: I heavily rely on GitHub Copilot to generate tests in my own workflows—but I still manually write a number of them to help formulate my thoughts.
In this article, I’ll walk you through why unit tests are essential, how GitHub Copilot can assist with generating unit tests, and practical tips for getting the most from Copilot’s test generation capabilities.
Oh, and if you’re curious I used Anthropic’s Claude model to generate the unit test examples you’ll find later in this article (in case you missed it, GitHub Copilot offers support for Anthropic’s Claude, Google’s Gemini, and OpenAI’s GPT o1 models).
Bringing developer choice to Copilot with Anthropic’s Claude 3.5 Sonnet, Google’s Gemini 1.5 Pro, and OpenAI’s o1-preview
At GitHub Universe, we announced Anthropic’s Claude 3.5 Sonnet, Google’s Gemini 1.5 Pro, and OpenAI’s o1-preview and o1-mini are coming to GitHub Copilot—bringing a new level of choice to every developer.
https://github.blog/news-insights/product-news/bringing-developer-choice-to-copilot/
Tomi Engdahl says:
Ocean Prompting Process: How to get the results you want from an LLM
https://www.raspberrypi.org/blog/ocean-prompting-process-how-to-get-the-results-you-want-from-an-llm/
Tomi Engdahl says:
Are Small Language Models (SLM) the Future of LLMs?
https://generativeai.pub/are-small-language-models-slm-the-future-of-llms-9fe202db8423
Tomi Engdahl says:
PRIME Intellect Releases INTELLECT-1 (Instruct + Base): The First 10B Parameter Language Model Collaboratively Trained Across the Globe
https://www.marktechpost.com/2024/11/29/prime-intellect-releases-intellect-1-instruct-base-the-first-10b-parameter-language-model-collaboratively-trained-across-the-globe/
Tomi Engdahl says:
https://www.theregister.com/2024/12/02/windows_xp_sweater_competition/
Tomi Engdahl says:
Out of the Lab and Into the Wild
WildLMa blends the latest in AI with VR to produce multitasking robots that might eventually give us a hand with our household chores.
https://www.hackster.io/news/out-of-the-lab-and-into-the-wild-a5ef6d184f91
Tomi Engdahl says:
https://blog.google/technology/research/behind-the-scenes-google-quantum-ai-lab/
Tomi Engdahl says:
https://artlist.io/
Get 10% off the Max plan
Sign in
All you need to create amazing videos
Get the highest-quality music, SFX, AI voiceovers, footage, and more.
Tomi Engdahl says:
https://parcero.fi/blogi/tekoaly-markkinoinnissa-markkinoijan-ai-tyokalupakki/
Tomi Engdahl says:
https://techxplore.com/news/2024-12-algorithm-robots-good-decisions-real.html
Tomi Engdahl says:
Make a Mark in Motion Capture Without Markers
The BioCV dataset is a treasure trove of synchronized video, motion capture, and force plate data for evaluating markerless capture systems.
https://www.hackster.io/news/make-a-mark-in-motion-capture-without-markers-e8b1d65c292f
Tomi Engdahl says:
Tim Berners-Lee wants the internet back You would think that a web increasingly driven by AI and AI content will be less open and free, but Berners-Lee is optimistic.
Read more at: https://www.deccanherald.com/opinion/tim-berners-lee-wants-the-internet-back-3282218
Tomi Engdahl says:
Tekoälyn vastuullisuuteen ja erityisesti energiankulutukseen on alettu kiinnittämään enemmän huomiota, mutta valitettavasti osa puhujista ajattelee asiasta kovin naiivisti – uskoen, että teknologian kehittyminen tulee ratkaisemaan energiankulutuksen. Tämä tuskin pitää paikkaansa, ellei tekoälyn vaatimassa laskennassa ja sen energiankulutuksessa tehdä jotain merkittävää kvanttihyppyä lähivuosina. Yllättävää kyllä, tekoälyn eettisyyden pohdintaa oli vähemmän kuin energiakeskustelua. Tässä ehkä näkyy teknologiakonferenssien tapa katsoa aina uusia asioita ja jättää aiemmin keskustellut pienemmälle huomiolle.
https://www.exove.com/fi/blogit/websummit-2024-tekoalyn-kytkeytyminen-kaikkeen/
Tomi Engdahl says:
Kasvavan energiankulutuksen lisäksi pohdinnassa on tekoälyn yleinen kestävyys, koska tekoälyratkaisut vaativat kiihtyvällä tahdilla dedikoitua rautaa.
Tomi Engdahl says:
Suositukseni
Tekoälyjunasta ei kannata jäädä. Tekoäly on tullut osaksi elämäämme ja sitä ei sieltä enää saada pois. Pärjätäkseen tulee – Aku Ankan taskukirjasta oppimani viisauden mukaan – ulvoa susien kanssa, mutta kovempaa. Eli nyt on hyvä (ja ehkäpä myös viimeinen) hetki laittaa tekoälystrategia kuntoon ja aloittaa tekoälyn testaaminen. Koska vaikka yrityksesi ei sitä testaisi, kilpailijasi kyllä testaavat.
Tekoälyn energiankulutus on vahvassa kasvussa. Tämä tuo mukanaan erilaisia epämiellyttäviä skenaarioita, joita vastaan on taisteltava sekä toimimalla fiksusti – valitsemalla tehtävään sopivin tekoälymoottori, pohdittava opetusta tarkasti ja rajattava tekoälyn käyttö arvontuottoa maksimoiviin toimiin.
Kuplan mahdollisuus kannattaa pitää mielessä riskienhallinnan kannalta. Mikäli tekoälyliiketoimintaan todellakin syntyy kupla ja yrityksen palvelut tulevat taholta, joka kärsii merkittävästi kuplasta, tekoälyratkaisut voi joutua siirtämään nopealla tahdilla toiselle toimijalle. Kannattaa miettiä munien jakamista eri koreihin tai sitoutumista tahoon, jolla on muitakin tulovirtoja kuin tekoäly.
https://www.exove.com/fi/blogit/websummit-2024-tekoalyn-kytkeytyminen-kaikkeen/
Tomi Engdahl says:
https://www.solita.fi/news/solita-ja-iss-tutkimushanke-miten-genai-muuttaa-ohjelmistokehityksen-tulevaisuutta/
https://www.solita.fi/news/tekoaly-suomalaisyrityksissa-tutkimus-suomalaiset-suuryritykset-hitaita-generatiivisen-tekoalyn-hyodyntamisessa/
https://www.solita.fi/services/generative-ai/?utm_source=google&utm_medium=cpc&utm_campaign=gen_ai&utm_content=service_site&utm_creative_format=text&utm_marketing_tactic=b2b&utm_term=generative%20ai&utm_region=finland&utm_campaign=B2B:+FI:+Traffic:+:+AI+%26+Generative+AI&utm_term=generative%20ai&utm_source=adwords&utm_medium=ppc&hsa_kw=generative%20ai&hsa_acc=2178929025&hsa_mt=b&hsa_grp=165831051031&hsa_tgt=kwd-545293778078&hsa_src=g&hsa_cam=21334044155&hsa_net=adwords&hsa_ad=714320967565&hsa_ver=3&gad_source=1&gclid=Cj0KCQiAx9q6BhCDARIsACwUxu46i1dFiucZUHUmow0WvNeaX_q19gzgpjo90pv3FDCXSmTk4Z1PuPIaAtVbEALw_wcB
Tomi Engdahl says:
https://www.sap.com/finland/products/artificial-intelligence/what-is-generative-ai.html
Tomi Engdahl says:
https://www.sttinfo.fi/tiedote/70205550/solita-ja-iss-tutkimushanke-miten-gen-ai-muuttaa-ohjelmistokehityksen-tulevaisuutta?publisherId=69819622&lang=fi
https://www.solita.fi/guides/suomen-top-500-yhtiot-genain-hyodyntajina/
Tomi Engdahl says:
Tutkimus: Suomalaiset etsivät tekoälystä työhönsä lisää tehoa ja mielekkyyttä – jo puolet toimistotyötä tekevistä hyödyntää generatiivista tekoälyä
https://www.sttinfo.fi/tiedote/70747923/tutkimus-suomalaiset-etsivat-tekoalysta-tyohonsa-lisaa-tehoa-ja-mielekkyytta-jo-puolet-toimistotyota-tekevista-hyodyntaa-generatiivista-tekoalya?publisherId=69819622&lang=fi
Tuoreen Tekoäly suomalaisessa työelämässä -tutkimuksen mukaan tekoälystä on tullut arkea jo lähes puolelle toimistotyötä tekevistä suomalaisista: Solitan tuoreen tutkimuksen mukaan 46 prosenttia suomalaisista toimistotyöntekijöistä käyttää generatiivista tekoälyä työssään. Suomalaiset hyödyntävät tekoälyä työssään selkeästi ruotsalaisia yleisemmin
Lähes puolet (46 %) suomalaisista toimistotyötä tekevistä kansalaisista käyttää generatiivista tekoälyä työssään. Kuusi prosenttia hyödyntää tekoälyä päivittäin, kun joka viides (20 %) hyödyntää tekoälyä viikottain.
Suurin motiivi generatiivisen tekoälyn hyödyntämiseen on tuottavuuden parantaminen: yli puolet vastaajista, (56 %), kertoo tekoälyn hyödyntämisen syyksi oman työn tuottavuuden parantamisen. Suomalaiset hyödyntävät tekoälyä työssään yleisesti myös oman tietämyksen ja ajattelun laajentamiseen (49 %) ja työn laadun parantamiseen (45 %). Rutiinitehtävien vähentäminen on motiivina tekoälyn käytölle 43 prosentille vastaajista. Saman verran vastaajista (43 %) haluaa pysyä ajan tasalla teknologian kehityksestä.
Joka viides hyödyntää tekoälyä päästäkseen vähemmällä
Lähes joka viides (19 %) hyödyntää tekoälyä työssään, jotta voisi työskennellä vähemmän. Vastaajista 8 prosenttia käyttää tekoälyä löytääkseen uusia liiketoimintamahdollisuuksia.
Yleisin syy sille, ettei tekoälyä käytetä omassa työssä on, ettei sen nähdä tuovan työhön lisäarvoa (43 %). Lähes yhtä suuri osa niistä, jotka eivät hyödynnä tekoälyä työssään, ei tiedä, mihin tekoälyä voisi omassa työssä hyödyntää (42 %). Viidennes vastanneista (18 %) kertoo puolestaan syyksi huolen tietoturvasta. Nuoret työntekijät olivat vanhempia työntekijöitä huolestuneempia tekoälyn turvallisuudesta.
Suomalaiset uskovat tekoälyn tuovan mielekkyyttä työhön
Tutkimus myös selvitti, miten suomalaistyöntekijät uskovat tekoälyn muuttavan omaa työtä seuraavan viiden vuoden aikana. Yli kolmannes vastaajista, 35 %, uskoo tekoälyn vapauttavan rutiinitehtävistä ja tarjoavan itselle mahdollisuuksia keskittyä mielekkäämpiin työtehtäviin. Vain 4 prosenttia vastaajista näkee, että tekoäly pystyisi suorittamaan suurimman osan tai kaikki omista työtehtävistä. Kuitenkin 43 prosenttia vastaajista näkee tekoälyn pystyvän lähivuosina suorittamaan osan omista nykyisistä työtehtävistä. Kolmannes vastaajista (33 %) uskoo tekoälyn muuttavan omien työtehtävien sisältöä.
Tomi Engdahl says:
Will AI Replace Programmers?
https://devot.team/blog/will-ai-replace-programmers
At the Bug Future Show 2024, our CEO, Martin Morava, sparked quite a conversation with his presentation titled “Farewell to Traditional Coders.” Sounds dramatic, doesn’t it? True to form, the bold statement raised a flurry of discussions in the comment section after the event.
The question of “AI replace software engineers” has been a hot topic as AI’s role in the IT industry continues to expand.
In this blog post, let’s go over how artificial intelligence is reshaping the landscape of software development and why the core of programming, thinking, and problem-solving means the future of the profession is safe
Today, AI technologies like generative AI and natural language processing have, in a way, changed the dialogue from whether AI can replace programmers to how it can augment their productivity and creativity.
Tomi Engdahl says:
What are the Advantages and Disadvantages of Using AI in Development?
https://beetroot.co/ai-ml/what-are-the-advantages-and-disadvantages-of-using-ai-in-development/
A year after the launch of ChatGPT, the matter of generative AI adoption still can cause heated discussions. However, despite different opinions on the technology, it’s hard to deny that embracing AI has become not an option but a necessity. And the software development field is no exception.
Artificial intelligence opens the door to unprecedented advancements in development practices. Even though the GenAI tools are relatively new, 2023 research by McKinsey showed that with generative AI-based tools, software developers could write new code nearly twice as fast and optimize existing code in nearly two-thirds of the time. As a software development company, Beetroot has also witnessed how AI empowers developers to complete projects faster, improve code quality, and tackle complex problem-solving scenarios.
Naturally, transformative technologies often come with certain challenges, and it’s crucial to acknowledge them. We prefer to refer to them as challenges, not disadvantages, because, like any worthwhile journey, the path to AI integration is not without obstacles. In this article, we explore how navigating these challenges can lead to a balanced and informed approach to leveraging AI’s potential in software development.
Benefits of AI in Software Development
According to Github’s survey, 92% of US-based developers already use AI coding tools in and outside of work. On top of that, 70% of the respondents said that AI coding tools provide them with a workplace advantage, citing better code quality, faster completion time, and effective incident resolution, among other benefits. Let’s figure out how these advantages of AI in software development work in practice.
Improved Accuracy and Bug Detection
Precision in software development directly impacts the reliability and functionality of the final product. Inaccuracies in code can lead to bugs, security vulnerabilities, system failures, and whatnot, jeopardizing the user experience and compromising data integrity.
One of the core benefits of AI in software development and tools for coding specifically lies in their ability to enhance accuracy by automating specific tasks and performing code analyses. For instance, static code analysis tools can meticulously examine code without execution and identify potential bugs and vulnerabilities before the software goes live. For this, developers often use tools like SonarQube and ESLint that offer real-time insights into code quality and alert about the issues that might have otherwise gone unnoticed.
Personalization and Advanced User Experience
The positive impact of AI tools in software development goes beyond coding. On top of making the product more reliable, AI has the power to significantly enhance user experience by enabling personalized and adaptive interfaces. Machine learning (ML) algorithms analyze user behavior, allowing developers to create adaptive interfaces that cater to individual preferences, providing smoother and more personalized interactions. As a result, this approach fosters increased user satisfaction and engagement.
Predictive Analysis for Better Decision-making
By leveraging predictive analysis, developers can foresee and address potential challenges before they escalate. In this case, AI tools can be especially helpful, contributing to software products’ overall reliability and performance.
Challenges of AI in Software Development
Nearly every transformative technology comes with nuanced challenges, and it’s essential to recognize and address them. That’s exactly what this section is about.
Complexity and Learning Curve
Incorporating AI tools can introduce a significant learning curve for software development teams. After all, beyond coding, mastering machine learning concepts and frameworks demands time and effort.
Let’s take a look at deep learning as an example. Adapting to deep learning techniques for natural language processing or computer vision requires developers to comprehend the fundamentals of neural networks. This learning curve can slow down the adoption of AI, necessitating extensive training programs and support to enable engineers to leverage these tools effectively.
Despite being a challenge, this learning curve is also a chance for growth. It emphasizes the importance of training and support for developers, acting as a bridge between coding and AI. As teams learn, they boost their skills, paving the way for smoother AI integration. Understanding these layers is critical to unlocking AI’s transformative potential in software development.
Data Dependency and Privacy Concerns
AI’s effectiveness relies on the availability and quality of data. However, obtaining diverse and relevant datasets comes with certain practical challenges that can impact the performance of AI algorithms. Securing access to substantial, diverse, and high-quality datasets is crucial for development teams. To deal with the challenge, developers need to employ rigorous data preprocessing techniques, which involve cleaning, filtering, and enhancing datasets to meet the standards of effective AI training. Moreover, the process may require collaborations and partnerships to access more extensive and varied datasets.
Integration with Existing Systems
Sometimes, legacy systems lack inherent compatibility with AI technologies, leading to potential obstacles during integration.
Ethical and Employment Concerns
Ethical concerns with artificial intelligence are one of the reasons the adoption of these tools faces resistance. The most significant of these issues are connected to algorithm bias and the potential impact of AI automation on employment.
Tomi Engdahl says:
What Are the 5 Limitations of AI in Low-Code App Development?
https://www.appbuilder.dev/blog/limitations-of-ai-in-low-code-development
What is low-code app development? How about AI in low code? What are the limits, challenges, and opportunities? Find out in this blog post.
The App Development Reinventor – Low-Code Platforms
What is low-code app development? In essence, it is the process of crafting full-featured apps for any framework using low-code platforms. To address many of the hurdles and bottlenecks that today’s app development teams face, these tools arrive with a set of capabilities that automate everything behind app building—from design to code.
Low-code development platforms offer many features designed to streamline the development process and enhance productivity. These platforms provide a user-friendly, drag-and-drop app building experience combined with high-speed rapid application development (RAD). They include a visual integrated development environment (IDE) for defining user interfaces, data models, workflows, and more, making it easier for developers to create complex apps without extensive manual coding.
A key advantage of low-code platforms is their comprehensive toolbox of reusable UI components for Angular, Blazor, and Web Components. Additionally, these platforms offer a design system that adheres to an inventory of UX patterns and brand style guidelines, ensuring that the applications function well and provide a consistent UX. Instant code generation and real-time code preview capabilities are also significant features of low-code platforms. These enable developers to see changes immediately, facilitating faster iteration.
AI Low Code: Understanding AI in Low Code
In summary, AI in low-code refers to the integration of artificial intelligence capabilities within low-code development platforms. This combination allows teams to leverage both technologies’ strengths, enhancing and speeding up the development process while reducing manual coding effort.
One example of the function of AI in low code development is AI-assisted programming, where AI analyzes user inputs and then generates code snippets, workflows, or entire apps. AI algorithms allow repetitive tasks to be automated and completed much faster than traditional hand coding.
With AI-powered natural language processing (NLP) capabilities, users can easily describe what they want in plain language and define parameters. The AI-powered platform then quickly scans the data and translates these descriptions into functional code or components. This way, people with a limited technical background can actually build applications easily and with little assistance.
However, despite these advantages and practical scenarios, there are certain disadvantages, as already mentioned.
Let’s Discuss the Limitations of AI in Low-Code Development
When we think about AI, we typically think of critical factors and results like providing new ways for:
Revolutionizing how tasks are automated.
Generating code outputs with AI code assistants
Improving business processes and solutions.
All this is thanks to its ability to understand and replicate patterns from vast datasets. Nevertheless, as the use of AI in app development continues to grow, it is not without certain limitations. For example, Gartner Survey reveals that “46% of CIOs are partnering with their CxO peers to bring IT and business area staff together to co-own digital delivery on an enterprise-wide scale”. While this democratization means saving time and development resources because the digital delivery of products and services is entrusted to non-technical teams, there are questions like what the technical consequences are that non-technical people can’t really foresee.
Here are other key challenges to consider:
Developers Losing Control Over Their End Product
While AI-driven features aim to simplify the development process, the main concern for developers is they do not want to lose control over what they are building. With tools like App Builder, the differentiator is that we give the end user complete control and editable code. Anything you produce inside the platform is yours, fully customizable with the extensibility to export production-ready code for Angular, Blazor, Web Components, or others.
Low-Quality, Non-Usable Code
The main advantage of low-code development tools is the ability to deliver production-ready, usable code. Here, we see another limitation in AI-generated code: It tends to focus on quantity rather than quality, bug-free code. This forces teams to sacrifice the long-term maintainability of their projects, hindering their ability to troubleshoot and debug any issues that arise quickly.
Flexibility and Adaptability
Low-code platforms provide users with the flexibility to tailor applications to their specific requirements, incorporating custom business logic, integrations, and workflows. AI models, on the other hand, may need significant training and customization to generalize across diverse use cases or adapt to evolving requirements.
User Experience and Control
While AI-driven features aim to simplify the development process, they struggle to balance automation and user control. Developers are pigeonholed to AI-generated suggestions that may not align with their vision or preferences despite describing certain parameters in the first place. This, in turn, removes the human element of app design, creating a lack of creativity and no longer fostering innovation.
Security and Compliance
To ensure that apps built with AI models comply with industry standards and requirements, meticulous code quality assurance is required, which automated processes may not be able to provide. As a result, this could lead to compliance issues and code errors or vulnerabilities, potentially exposing the entire project to security risks and legal complications that are costly and time-consuming to resolve post-deployment.
Tomi Engdahl says:
https://binmile.com/blog/pros-and-cons-of-ai-assisted-coding/
Tomi Engdahl says:
Common Pitfalls of Using AI Tools in Software Development. How to Avoid Traps
https://codeandpepper.com/common-pitfalls-ai-tools-software-development/
In the evolving landscape of software development, artificial intelligence (AI) tools have emerged as indispensable assets, promising efficiency, accuracy, and innovation. However, like any powerful technology, AI tools come with their own set of pitfalls. Software developers, while leveraging these tools, must navigate several traps to harness AI’s full potential without compromising the quality, security, and ethics of their projects.
Over-Reliance on AI Tools
One of the primary traps is over-reliance on AI tools. While AI can automate many tasks, from code generation to bug detection, developers might fall into the trap of depending too heavily on these tools. This dependency can lead to a deterioration in fundamental programming skills and complacency, where developers might not question the outputs provided by AI, assuming them to be infallible.
“Initially, I wrote the code myself, knowing exactly what I wanted to write. Only when CodeWhisperer started suggesting code after a few lines did I realize I was falling into a trap. The time between writing my code and waiting for a suggestion, even if less than a second, was the time I spent thinking about what it might suggest. I was wasting time instead of writing myself immediately. AI can be a trap for programmers in this case because, over time, if we rely on such solutions more frequently, we might become lazy and not work on remembering the code we create. In my opinion, this is a threat,” says Wawer.
Misunderstanding AI Limitations
AI tools, despite their advanced capabilities, have limitations. They are trained on specific datasets and designed to perform particular tasks. Misunderstanding these limitations can lead developers to use AI tools inappropriately.
A few months ago, another Code&Pepper developer, Łukasz Duda, was testing Amazon CodeGuru Security:
“Everything looks good on paper and in theory. Since we use JavaScript at Code&Pepper, I wanted to test it. Using the test code prepared by AWS, I ran it through the AWS tool, and it did not find any security errors. Then I performed similar actions for Python and received interesting results. I tested the code written in JavaScript on another tool, Snyk, and the report showed a lot of errors. I repeated the test several times with the AWS tool and consistently received the same report, which did not detect any errors. The conclusion is that it works for Java and Python, but there is a long way to go for it to work for JavaScript,” says Duda.
Tomi Engdahl says:
AI tools in software development: risks and benefits
https://www.llinformatics.com/blog/ai-tools-in-software-development
Benefits of AI coding assistants
One of the most significant benefits of AI-charged code generation is a boost in productivity. These tools can automate repetitive and mundane tasks, such as code formatting, syntax corrections, and basic debugging.
By handling these routine activities, AI development assistants free up developers to focus on more complex and creative aspects of their work. This leads to faster development cycles and allows teams to deliver projects more swiftly. Furthermore, it reduces the risk of human error in these routine operations.
Advanced code search and navigation
Navigating large codebases can be challenging, especially when trying to understand legacy code or track down specific functionalities. AI coding tools offer advanced code search and navigation capabilities, allowing developers to quickly locate relevant sections of code, understand dependencies, and trace the flow of execution. This is particularly beneficial in large-scale projects where understanding the entire codebase is impractical for any single developer.
Smart code completion and suggestions
While traditional IDEs offer basic code completion, AI development assistants take it a step further with context-aware suggestions. These tools analyze the current context of the code, previous patterns, and even the developer’s coding style to provide more accurate and useful suggestions. This can significantly speed up coding by reducing the amount of time spent typing and correcting code. However, it’s important to note that these suggestions are not infallible and can be incorrect approximately 32% of the time, requiring developers to verify the AI-generated code.
Seamless integration with development tools
AI development assistants integrate seamlessly with popular development environments and tools, such as Visual Studio Code, IntelliJ IDEA, and GitHub. This integration ensures developers can leverage AI capabilities without disrupting their existing workflows. For example, AI can assist in version control by suggesting commit messages, identifying potential merge conflicts, and recommending resolution strategies.
Continuous learning and improvement
AI coding tools continuously learn from the developer’s interactions and the code they work on. This ongoing learning process allows the AI to improve its suggestions and adapt to the specific needs and preferences of the development team. Over time, this results in more accurate and contextually relevant assistance, although it still requires senior software developer oversight to ensure reliability.
The biggest challenges of AI tools for code generation
AI development assistants are transforming the landscape of software engineering. Their ability to enhance productivity, save time on repetitive tasks, provide advanced code navigation, and offer smart code completion makes them indispensable tools for modern development teams. However, it is crucial to remain aware of their limitations. A 2023 study by researchers at Bilkent University found that AI assistants produced code with a 30.5% error rate, while an additional 23.2% of the code was only partially correct. The accuracy rates varied among different AI code generators and here’s how it can affect your project.
Unfixable code written by AI
AI-charged code often lacks the intuition and contextual awareness of human developers. This results in code that, while syntactically correct, can be semantically puzzling. Developers tasked with maintaining such code frequently face significant challenges. The AI’s approach to solving a problem might involve unconventional patterns or algorithms that aren’t well-documented or commonly understood, leading to difficulties in debugging and extending the code.
Lack of documentation and rationale behind AI-generated solutions
AI tools typically don’t provide comprehensive documentation or explanations for their code. This absence of rationale makes it hard for developers to grasp why certain decisions were made, complicating efforts to modify or enhance the code. Without proper documentation, the code becomes a black box, and developers must spend considerable time reverse-engineering it to understand its functionality and purpose.
High error rate of AI assistants
AI coding assistants are prone to errors, such as syntax mistakes, logic flaws, and improper use of libraries or APIs. These errors stem from the AI’s limited understanding of context and the specific nuances of the codebase it is working on. Codesignal’s poll showed that 55% of developers have concerns about the quality of AI code and 48% voice concerns about security and/or privacy.
Impact of a 30.5% error rate on project timelines and costs
A 30.5% error rate can have a profound impact on project timelines and costs. For every ten lines of code generated, three may require significant rework, delaying development schedules and increasing labor costs. The time spent identifying and correcting these errors can negate the productivity gains from using AI tools, making them less appealing for critical projects.
Overwritten code and complex architectures
One major risk with AI tools is their potential to overwrite existing, functional code. This can occur when the AI misinterprets the requirements or attempts to optimize code without understanding its current functionality. Such overwrites can introduce new bugs and regressions, disrupting the stability of the application.
Challenges in integrating AI-generated code with existing architectures
Integrating AI code into existing architectures can be fraught with difficulties. AI coding might not adhere to the architectural patterns and design principles already in place, leading to conflicts and inconsistencies. This can result in a disjointed codebase that is harder to maintain and scale.
The complexity added by AI solutions and its impact on project maintainability
AI assistant’s code might involve advanced algorithms or patterns that are not well understood by the team, making future modifications challenging. This added complexity can increase the technical debt of a project, complicating maintenance and reducing overall code quality.
Lack of flexibility, maintainability, and scalability
AI-generated code tends to be rigid and inflexible, making it difficult to adapt to changing requirements. Unlike human developers who can anticipate future needs and design code with extensibility in mind, AI coding often produces solutions that are narrowly focused on the initial problem statement.
Maintenance challenges posed by AI-created solutions
Maintaining AI-charged code can be more challenging than maintaining human-written code. The lack of context and documentation, combined with potential unconventional coding practices, makes it harder for developers to understand and modify the code. This can lead to increased maintenance costs and longer turnaround times for bug fixes and enhancements.
Scalability issues due to non-optimized AI code structures
AI-charged code might not be optimized for performance and scalability. The AI assistants’s lack of understanding of the broader system architecture can result in solutions that do not scale well under increased load or that perform poorly in production environments. Addressing these issues often requires significant refactoring and optimization by experienced developers.
Security concerns from a business standpoint
AI tools can inadvertently introduce security vulnerabilities into the codebase. These vulnerabilities might arise from improper handling of user inputs, incorrect implementation of security protocols, or the use of insecure libraries. Such vulnerabilities can be exploited by attackers, posing significant risks to the application and its users.
Risks of relying on AI for security
Relying on AI for security-critical components is particularly risky. The AI might not fully understand the intricacies of security requirements, leading to implementations that are susceptible to attacks. Human oversight is crucial in these scenarios to ensure that security measures are correctly and thoroughly applied.
Business implications of security vulnerabilities
Security breaches resulting from AI-generated code can have severe business implications. These include financial losses, reputational damage, legal liabilities, and regulatory penalties. Ensuring the security of AI coding assistant’s code is therefore paramount to protecting the business and its stakeholders.
Dependence on AI
Over-reliance on AI coding can lead to a decline in developer skills. If developers become too dependent on AI for routine coding tasks, they may not develop the problem-solving skills and deep understanding of the codebase necessary for more complex tasks. This skill erosion can be detrimental to the team’s overall effectiveness and innovation capacity.
Risk of AI coding without proper human oversight
AI-charged code without proper human oversight can lead to a host of issues, from functional errors to security vulnerabilities. Ensuring that human developers review and refine AI-generated code is essential to maintaining code quality and project integrity.
Security and Privacy Issues
AI tools themselves can have vulnerabilities that, if exploited, could compromise the security of the entire development environment. These vulnerabilities might include insecure APIs, improper access controls, and insufficient protection of sensitive data.
Data Privacy Concerns with AI Tools Processing Sensitive Information
AI tools often process large amounts of data, including sensitive information. Ensuring that these tools comply with data privacy regulations and standards is critical to preventing data breaches and protecting user privacy. This involves implementing strong data encryption, access controls, and audit mechanisms to safeguard sensitive information.
Limitations of AI assistant tools
Despite their advanced capabilities, AI coding assistants have several limitations:
Context limitations: AI coding assistants might not fully grasp the broader context of a project, leading to code suggestions that are technically correct but contextually inappropriate.
Creativity and innovation: AI assistants can generate code based on existing patterns and best practices, but they might struggle with highly innovative or creative solutions that deviate from common patterns.
Specialized knowledge: They may not have deep expertise in niche domains or specialized fields, where human experts with specific knowledge are necessary.
Dependency on training data: The effectiveness of AI coding assistants depends on the quality and diversity of the training data. If the training data lacks examples of specific use cases, the assistant’s code suggestions may be less relevant.
Ethical and security considerations: AI-charged code can sometimes include insecure or unethical practices if such patterns were present in the training data. Human oversight is necessary to ensure code quality and adherence to ethical standards.
Bias and errors: AI models can exhibit biases based on their training data, leading to code suggestions that reflect those biases. Additionally, they might generate errors or suboptimal code, requiring careful review by developers.
Learning and adaptation: While some AI assistants can learn and adapt to user preferences, this process may take time and may not always align perfectly with the user’s specific needs or coding style.
Integration challenges: Integrating AI coding assistants into existing workflows and toolchains can sometimes be challenging, especially in highly customized or proprietary.
Tomi Engdahl says:
AI in software development: Key opportunities + challenges
AI in software development will change how engineers design, develop, and deploy products. Learn how AI will transform software engineering in the 21st century.
https://www.pluralsight.com/resources/blog/business-and-leadership/AI-in-software-development
Tomi Engdahl says:
The Limits of AI Assisted Software Development
https://medium.com/@earlred/the-limits-of-ai-assisted-software-development-8e3467023f0e
What are the limits of AI coding?
The use of artificial intelligence (AI) in software development has been growing rapidly.
While AI-assisted software development offers significant advantages, it is unlikely to completely replace human developers.
One notable limitation is that current models still struggle to generate functional code beyond simple snippets.
There are concerns regarding security vulnerabilities associated with using AI-generated code.
Legal concerns related to copyright and intellectual property when it comes to AI-generated code exist.
Humans possess unique capabilities and perspectives that cannot be replicated by AI systems.
Tomi Engdahl says:
https://claude.ai/login?returnTo=%2F%3Futm_source%3Dgoogle%26utm_campaign%3D21532474528%26utm_medium%3Dcpc%26utm_content%3D707830711165%26utm_term%3Dbest%2Bai%2Bfor%2Bcode%26targetid%3Dkwd-2073843249317%26gad_source%3D1%26gclid%3DCj0KCQjwmt24BhDPARIsAJFYKk0gNHF1BrB5mHxV_HJN_Nily8mFI5Ds9E7GX-HH3YsWXL-XF2cJo60aAnZhEALw_wcB
Tomi Engdahl says:
The AI Code Editor
https://www.cursor.com/
Built to make you extraordinarily productive,
Cursor is the best way to code with AI.
Tomi Engdahl says:
https://aws.amazon.com/q/developer/?gclid=EAIaIQobChMIrf_V8amqigMVbWSRBR0BRDIvEAAYAiAAEgKDBvD_BwE&trk=45c2a23c-23fc-4f2e-b28a-4d8b6ef62ce7&sc_channel=ps&ef_id=EAIaIQobChMIrf_V8amqigMVbWSRBR0BRDIvEAAYAiAAEgKDBvD_BwE:G:s&s_kwcid=AL!4422!3!698133085581!p!!g!!ai%20for%20coding!21048268977!166963732772
Tomi Engdahl says:
https://fastflux.ai/
We’ve generated over 500 million images through FastFLUX since its launch, and we’re excited to announce that the demo has now moved to our dashboard with even more features and models. You can continue creating amazing images for your projects at the lowest cost in the industry.
As a thank you for your continued support, new users who register on our platform with a business email will receive ~1000 free images to get started
Tomi Engdahl says:
https://gemini.google.com/
Tomi Engdahl says:
Mikä on generatiivinen tekoäly?
Voit keskustella generatiivisen tekoälyn kanssa mutta tiedätkö miten se toimii? Tämä selviää pian!
https://www.finnishup.com/mika-on-generatiivinen-ai/
Generatiivinen tekoäly, joka tunnetaan myös nimellä GenAI, on edistynyt tekoälyn ala, joka keskittyy algoritmien luomiseen, jotka pystyvät tuottamaan uutta sisältöä. Se voi tarkoittaa kaikkea kuvista, tekstistä, musiikista tai jopa kolmiulotteisista malleista.
Generatiivisen tekoälyn periaatteet
Generatiivisen AI:n taustalla on usein syvät neuroverkot joita kehitetään aktiivisesti uuden datan, koulutusmallien ja optimintitekniikoiden avulla.
Generatiivisen tekoälyn 4 vaihetta
Generatiivisen tekoälyn voi nähdä omaksuvan neljä vaihetta:
Vaihe 1: Koulutus
Ensimmäisessä vaiheessa luodaan generatiivisen tekoälyn neuroverkko, joka ottaa lähdetietokannaksi esimerkiksi kuvia, videoita, ääniä tai tekstiä.
Vaihe 2: Kehoitus tai kysely
Useiden generatiivisen tekoälyn sovellusten kuten ChatGPT:n ominaispiirre on että siltä voi saada erillaista tietoa syöttämällä kehoitteen (englanniksi prompt) tai kyselyn saadakseen tekoälyn käyttöliittymästä tietoa.
Kehoitus viittaa siihen, kun tekoälyä ohjataan tai “kehotetaan” tuottamaan tiettyä tyyppiä olevaa sisältöä antamalla sille alkusysäys tai esimerkki, kuten aloituslause tekstissä, joka määrittää sen, minkälaista sisältöä AI:n tulisi jatkaa tai generoida. Esimerkiksi tekstipohjaisessa generatiivisessa mallissa kehoitus voi ohjata tekoälyn kirjoittamaan tietystä aiheesta tai tietyllä tyylillä.
Esimerkki kehoituksesta jonka voit syöttää ChatGPT:hen:”Olet Elias Lönnrot. Luo lyhyt runo siitä kuinka tekoäly korvaa sammon Kalevalassa.”
Vaihe 3: Tekoäly luo sisällön
Kolmas tärkeä vaihe generatiivisessa tekoälyssä on kun tekoäly luo sisällön vastatakseen käyttäjän kehoitukseen. Vastaus jälleen hyödyntää neuroverkkoja luodakseen täysin uutta sisältöä esimerkiksi tekstin, äänen tai videon muodossa. Vastaus perustuu koulutusdatan tulkitsemiseen ja voi siis olla erillainen riippuen siitä kuka luo kehoituksen ja onko samanlaisia kehoituksia tehty aikaisemmin.
Vaihe 4: Tekoälyn jatkokehitys
Useimmat suuria kielimallia hyödyntävät tekoälysovellukset, kuten ChatGPT, kehittyvät jatkuvasti sen mukaan kun niille syöttää enemmän koulutusdataa. Neljäs askel generatiivisen tekoälyn jatkokehityksessä on jatkuva kehitys, joka näkyy käyttäjän silmissä uusina kielimalleina kuten GPT-3 tai GPT-4 jne.
Generatiivisen AI:n sovellukset
On lukemattomia eri tapoja joilla generatiivista tekoälyä voi hyödyntää niin arjessa kuin työelämässä. Tässä muutamia yleisiä sovelluksia:
Kuvan luominen ja muokkaus: Generatiivista AI:ta voidaan käyttää luomaan realistisia kuvia tai muokkaamaan olemassa olevia kuvia. Esimerkiksi, se voi muuntaa sateisen kuvan aurinkoiseksi päiväksi tai luoda ihmisen kasvokuvan, joka ei perustu todelliseen henkilöön.
Tekstin tuottaminen: AI voi luoda kokonaisia artikkeleita, runoja tai tarinoita. Generatiivinen AI voi myös optimoida markkinointitekstiä tai tuottaa uutisotsikoita.
Musiikin, puheen ja äänen tuotanto: Generatiivinen tekoäly voi säveltää musiikkia tai luoda realistisia ääniä. Esimerkiksi se voi tuottaa taustamusiikkia videolle tai luoda ääniefektejä elokuviin.
Chatbotit: Generatiivisella tekoälyllä voi luoda interaktiivisia keskustelubotteja esimerkiksi asiakaspalveluun, tai yksinkertaisesti chat-assistenteiksi arjen jokapäiväisissä haasteissa ja kysymyksissä.
3D-mallinnus: Tekoäly voi luoda kolmiulotteisia malleja esineistä tai maisemista, joita voidaan käyttää virtuaalisessa todellisuudessa tai videopelien kehityksessä.
Tomi Engdahl says:
https://chat-gpt-suomi.fi/gpt-kysymys/voiko-chatgpt-kirjoittaa-arduino-koodia/
Tomi Engdahl says:
How To Run ChatGPT on your Arduino Projects
https://www.youtube.com/watch?app=desktop&v=EAwh4ul-K0g
How to integrate ChatGPT, a powerful language model, with your Arduino projects using a ESP32. Learn the steps to set up ChatGPT on Arduino, including hardware requirements, software dependencies, and libraries. Bring interactivity to your creative endeavors.
Tomi Engdahl says:
https://copilot.microsoft.com/?showconv=1
Tomi Engdahl says:
https://support.google.com/gemini/answer/15216790?hl=fi
Generatiivinen tekoäly voi auttaa sinua oppimaan itsellesi sopivalla tavalla. Kurssimateriaalien tarkistamisen lisäksi generatiivinen tekoäly voi auttaa sinua valmistautumaan kokeisiin muillakin tavoilla:
Monimutkaisten käsitteiden ymmärtäminen luovilla tavoilla: Esimerkiksi: “Kirjoita räppi, jonka avulla ymmärrän trigonometrisiä suhteita.”
Vaiheittaiset ohjeet: Voit esimerkiksi pyytää yksityiskohtaisia ohjeita matemaattisten ongelmien ratkaisemiseen.
Tietojen testaaminen: Pyydä niin monta harjoituskysymystehtävää kuin tarvitset ollaksesi valmis kokeeseen.
Analysoi tietoa nopeasti
Voit analysoida tietoa nopeasti generatiivisen tekoälyn avulla. Se voi auttaa esimerkiksi näillä tavoilla:
Tärkeiden asioiden selvittäminen: Se voi esimerkiksi auttaa sinua tutkimaan korkeakouluja tai tutustumaan eri pääaineisiin. Voit pyytää sitä esimerkiksi tekemään listan korkeakoulujen pääaineista eläimiä rakastaville ihmisille.
Erilaisten näkökulmien ymmärtäminen: Voit esimerkiksi valmistautua väittelyyn generatiivisen tekoälyn avulla pyytämällä sitä väittelemään vastapuolen puolesta.
https://support.google.com/gemini/answer/13954172?hl=fi