Microsoft ramps up A.I. game with bet on startup that helps coding novices build apps
Source: CNBC
LONDON Microsoft invested an undisclosed sum into Builder.ai, a startup that helps companies make applications without any coding experience, doubling down on its artificial intelligence efforts.
Founded in 2017 and headquartered in London, Builder.ai falls into the camp of startups that make so-called no-code and low-code platforms. Its software allows anyone from tech-shy artists looking to sell their work online to design professionals with limited programming experience to develop and manage apps.
Through a strategic partnership, Microsoft plans to integrate Builder.ais own AI assistant Natasha into its Teams video and chat software to let customers build business apps within the platform. Builder.ai will also enhance Natasha by leveraging Microsofts AI algorithms to make it sound more human, the company said.
-snip-
Were all convinced that the future of software is going to be where the customer doesnt need to be technical, Duggal told CNBC in an interview. What were really doing is bringing together a world where customers are able to build software, run software, host software.
-snip-
Read more: https://www.cnbc.com/2023/05/10/microsoft-ramps-up-ai-game-with-bet-on-no-code-startup-builderai.html
Learn to code, they said...
tinrobot
(12,050 posts)Glad Zuck or Elon didn't grab it.
Microsoft's stock is up today. Good decision for them.
Oneironaut
(6,288 posts)Microsoft has a tendency to make their UIs utterly excruciating when trying to make them friendly to new users. Its kind of like, I know youre trying to access something, but, were assumed youre a complete novice, so, well make it utterly hopeless to actually get there.
highplainsdem
(61,888 posts)To make them unnecessary.
Oneironaut
(6,288 posts)mbusby
(825 posts)of being a software engineer for 45 years before retiring .. just in time.
tinrobot
(12,050 posts)It's not fun, which is why most tech companies already outsource a lot of that work.
You don't want an engineer making 6 figures wasting time typing out "if/else" statements.
Oneironaut
(6,288 posts)In high level languages a lot is watered down and simplified anyways. I personally found writing code / software dev / software engineering to be more about trying to design things well with user experience in mind than drudgery.
99% if the time Ive spend on apps is designing, tweaking based on user feedback, and, debugging. It would be pretty cool if AI could help, with those things, but, it would need to be really smart and savvy beyond a point it is at right now.
honest.abe
(9,238 posts)I am an experienced programmer and I have also worked quite a bit with AI programming systems and I feel zero threat from AI programing bots. In fact I feel AI tools will actually make me more productive and more valuable. There might be some effect on low level junior programmers or outsourced programming work but it wont replace "most programmers".
Oneironaut
(6,288 posts)I would be surprised if AI completely replaces developers in the next 30 years, even. I think your last sentence is extremely accurate - outsourced junior .NET roles might be affected, but, other things require more nuance.
I can see AI being at best a coding aid for a long while. For example, it would be really nice to have AI show you code examples for different problems youre having, like you see now from Stack Overflow and stuff like that. Thats a far cry from taking over the entire development pipeline though. Lol
Yavin4
(37,182 posts)Last edited Thu May 11, 2023, 06:05 PM - Edit history (1)
Programming (writing code) is like being able to play the piano really well. Software Engineering is being able to compose a symphony.
Excellent analogy.
NullTuples
(6,017 posts)Meanwhile, with AI much of the output is necessarily garbage and unless things change drastically always will be, the output is not legally eligible for trademark due to not being made by a human, and at best AI will become - in my opinion - no more than a tool for people who still need to know what they are doing.
Will companies use it to make cheap copy or graphics to avoid hiring an artist or writer? Sure, but they would've avoided hiring one anyway and used a non-qualified person to create equally bad output.
AI has already been removing jobs in areas such as call centers and underpaid interns, but nobody much cared.
Finally, if you want to get to the root of all of this "gotta make it even cheaper", just look to private equity companies taking over anything everything with a popular brand that has any financial vulnerability and then cheapening the product while they extract all value to give to their investors. Without them, creative AI would barely even have a market.
Oneironaut
(6,288 posts)People chase trends to an annoying extent and get a little over-excited. Im glad the ChatGPT!!1! obsession is starting to die down.
Yes, both will probably add challenges and improvements to our lives weve never seen before. However, I always love the combinations of doomsdayers and overly eager futurists telling us how In 5 years well all be cyborgs!!1! and crap like that. Lol
Prairie_Seagull
(4,662 posts)eggplant
(4,181 posts)I see "no-code" AI systems to be most effective as fancy Siri/Google/Alexa engines, where you just tell it what you want and it does it, with ever increasing task complexity. The irony here is that you still have to tell it what you want each time. AI could, over time, learn your predilections and make this process simpler (assuming you tend to want the same thing each time). But AI will be replacing the programming aspect, not assisting with humans with writing code.
If I go into a restaurant and order a meal, I don't want the chef to show me how to make it. I just want to eat it.
tl,dr; "a world where customers are able to build software" will never happen. A world where customers can just ask for what they want and be given it is much more likely.
eggplant
(4,181 posts)That said, it is hubris to believe that AI-based coding won't be able to replace human coding at ever increasing levels of complexity. We're right on that cusp today. As AI learns to improve its results, it will quickly learn to build systems that mere humans can't comprehend, yet still accomplish the given task.