The First Software Jobs AI Will Replace Are...
HTML-код
- Опубликовано: 4 май 2024
- What will be the first software jobs replaced by AI (artificial intelligence)? In this clip, Dave is speaking to Birgitta Böckeler about what the future might hold for software engineer jobs in the face of emerging AI technologies. Birgitta is the lead at Thoughtworks for AI-assisted software delivery and has been working with these new technologies and forecasting what it might mean for our industry.
You can listen to the FULL episode on Spotify here ➡️ open.spotify.com/episode/2gbW...
-
🗣️ THE ENGINEERING ROOM PODCAST:
Apple - apple.co/43s2e0h
Spotify - spoti.fi/3VqZVIV
Amazon - amzn.to/43nkkRl
Audible - bit.ly/TERaudible
-
🙏The Engineering Room series is SPONSORED BY EQUAL EXPERTS
Equal Experts is a product software development consultancy with a network of over 1,000 experienced technology consultants globally. They increase the pace of innovation by using modern software engineering practices that embrace Continuous Delivery, Security, and Operability from the outset ➡️ bit.ly/3ASy8n0
-
#softwareengineer #softwarearchitecture 
Наука
Data Analysts, Transportation and Logistics, Manufacturing,Customer Service, Healthcare Diagnostics, Financial Services, Legal and Compliance
Basically very constrained systems, and searching in a database and doing pattern matching.
Constrained: data analysis, transportation, manufacturing, financial services
Search: customer service, healthcare diagnostics, legal and compliance
Very simple legal documents can be generated using AI, but not any contract I would use in a business I own or manage. For example, a typical non-disclosure agreement is two to three pages. The non-disclosure agreement I wrote is seven pages plus a signature page. I write contracts with the same level of detail, solid logic, and "error handling" to retain most of the agreement if any one provision is held to be unenforceable. My NDA was submitted for review by one of the top law firms in the region and I received only the comment that the agreement would not be binding without a quid pro quo, a fact I was already aware of.
People who accept mediocrity in all things are placing too much confidence in generative AI. The current state of AI limits it from replacing all but menial tasks like customer service.
As for manufacturing, that industry has been using increasingly sophisticated robotic controls for decades to automate discrete processes and to maximize the efficient use of material, equipment, and labor to ensure the right products are made at the right time to meet customer demand. Likewise, the transportation and logistics industry has also been using algorithms which optimize efficiency based on cost and schedule constraints. Generic AI models are not going to be more efficient than dedicated software applications which have been refined over the past 40+ years.
There are legal issues yet to be considered about using AI to replace healthcare diagnostics, financial services, and legal/compliance. No underwriter is going to issue professional liability coverage for even the most advanced AI currently available, and no business would put it to any of these uses without such insurance.
The first software jobs to be replaced will be tech recruiters. Even the real, live ones don’t know what they’re talking about, make snap decisions based on wrong assumptions and just spend most of their time employing sales tactics. If they behave in a manner indistinguishable from bots, I don’t see why not.
You're onto something.
2:16 The UI is really just a projection/transformation of data in flight and the backend is really just a transformation of data in flight to persisted data. Both well established repeatable patterns. The hardest part therefore is modelling the data in flight. The transformations to persisted data store or UI is just repeating the same patterns.
I guess we need a broad understanding of how the algorithms in AI work, especially their advantages and disadvantages. A black box like ChatGTP is a lost battle regarding testing and verification. You can make ever bigger models, tweak bias and stuff and you can't be sure where the black box breaks. And still you don't have the time and resources to test and verify the black box.
I'd be happy if the tools wouldn't cause me unnecessary work. It's yet not comparable to delegating certain task. There are other more reliable algorithms for simple work flows.
For me that translates into reading about the math and theory. My employer expects reliable software rather than a scientific paper/ experiment.
6:27 LLMs will never replace programmers because they are missing two important features. They don’t understand the actual AST of a program and they cannot make manipulations, transformations, additions, subtractions of that AST. When I say AST I’m not meaning the actual AST of single language, but the entire generalised concept of what an AST would be for a multi-language program like we see in web apps with SQL, backend code, over the wire serialisations, HTML, JS, etc, etc.
"we will never fly to the moon or have more than 640kb in our computers"
@@HoD999x LLMs aren’t the only AI techniques. Maybe a sufficiently complex LLM could gain enough emergent properties to do it. However, I’m placing my money on different techniques that do the things mentioned in my comment more efficiently, simply, and cost effectively.
Separation of concerns already exists, not everybody's a full stack developer. I don't see why the programming paradigms would stay static when given the ability to have specialist agents perform the tasks for far less money.
Dave is obviously a backend engineer
It will only replace mediocre people. LLMs by. design, produce mediocre text. Their weights are optimized for the average of the training data, which minimizes their loss function during training.
Both of them don't know how LLMs / "AI" work.
...the tedious ones.
Wrong way to look at it. AI does a good job when it was trained with a lot of examples. So, right now, AI does a good job with PHP, Ruby, Python, Javascript. If you expect good results with Scala, Julia or other obscure or less used languages, you’ll fail. Java also has a lot of code out there too, but maybe not as much in the open.
But it’ll be a while before AI can deliver more than functions which you clearly describe.
I use Copilot with Scala, and it works how I'd expect it to: autocompletions that follow the codebase style or in a functional style. I think LLMs will still be able to generalize no matter the source language.
@ Maybe Smalltalk training datasets help in that case.
I use Copilot with both Java and Clojure. The Clojure is output is definitely worse - usually not as I would want it. Even for Java, it tends to have bad variable names, so I have to go through and rename things.
@@bobbycrosby9765 We get used to magic really fast. Soon we will say: Why do I have to push that button to generate my project?
.. interns.
AI don't write unit tests.
Must have learned that from me!
@@lodrnr ha ha
They can write if they are told to write.
@@danilomenoli thats a recipe for reward hacking
neither do humans....