What is dead and/or going to die, and what is not dead yet

Having now experimented with LLMs and other things like GANs extensively I've come to the following conclusions:

The following are going to be replaced / mostly AI-done within a few years:

- Reading books - by most people (what's the point? You can just upload the PDF and summarise it). Most people can barely bring themselves to read books before AI, nevermind now that you can just get a summary.

- Writing books unaided - especially classic books - without adulteration by AI content

- Music generation - what's the point now that we have the likes of Udio?

- Stock images and photography - as soon as things like Midjourney produce reasonable hands. At the moment I've already given up on stock photos.

- Art prints - anything you want you can describe and print, even things that look like paintings.

- Video creation/ movies - so far we're up to 10 minutes. That's within a couple of years of the technology reaching the public. I estimate that by the end of 2026 we'll have feature films. Youtube content at the time of writing in my own feed (hand selected topics) - is about 10% AI at the moment.

- Research reports - at the moment you still have to help the AI but once you have adversarial agents it will become pointless, except perhaps for prompting novel ideas, I think humans are still good at that.

- Psychotherapists and motivational / life coaches. For basic consultation, even MDs can be replaced. Dr Google can step aside. We've even seen AIs perform surgery via robotic arm.

- Teaching, or at least, formal assessment (marking), and theory classes. Plain theory subjects like language learning and history can easily be made 100% digital, indeed, most people nowadays learn language through apps already.

- Fintech sector such as insurance assessment, digital fraud detection, banking anything, etc.

What won't be replaced YET:

- Live performances of music, but we have already seen this ten years ago: https://www.youtube.com/watch?v=fVUvWOESZLY 

- Physical paintings with impasto textures or similar, but 3D printing will be able to do it soon. In general, hand-crafted artworks that are actual paintings or sculptures. As budding or potential artists give up in the face of the onslaught of AI and 3D printing, those left standing will be in demand for their craft. The analogy I always use is shoes from China vs Italian hand-crafted. The wealthy will still buy italian.

- Psychotherapists / Life coaches / Medical doctors / general practitioners / specialists. The personal touch and in-person access is still important.

- Programmers. We still need humans to build the app, copy/paste the suggested code, open the right app, etc. However once we allow scripted agents to control our computers, that will be a thing of the past as well... with the accompanying security risks.

- Teaching in person, specifically lessons like physical education, chemistry labs, etc. Possibly art, cooking and other practical skills may survive. 

Popular posts from this blog

Recent Experiments testing ChatGPT's limits and some useful prompts

Introduction ChatGPT has a lot of people excited, but it deserves this level of excitement. It has a number of faults which I list below - some with workarounds - but in general it is amazingly good. It is absolutely not for having fun chatting to a robot, it's really more like a brilliant coworker who is incredibly stupid at some quite mundane tasks, but incredibly good at more tricky tasks. Summary of findings so far.  1. API vs Interactive Version - a caution 2. Coding Strengths and Weaknesses - be very careful but very useful anyway 3. Shell Scripting - amazingly good, probably what it's best at 4. Text Summarisation - fiddly but useful 5. Spreadsheet Analysis - annoying, maybe do by hand 6. Multi-Step Instruction Handling - very annoying 7. Spellchecking and writing style/grammar - ok overall but fiddly 8. File Compatibility and Formatting - not great 9. APA References from DOIs - good 10. Reference Accuracy - bad, double-check its work 11. Tone and Style Adaptations - ex...
Read more

Testing ChatGPT on coding - a script to make CSV to ICS

 So I gave ChatGPT a task of writing a CSV to ICS convertor, to take a spreadsheet of events and turn it into individual calendar file invitations. The idea is you can make a calendar with lots of event items in a spreadsheet, and then output ICS that you can just double-click to add to your calendar automatically. So here's the spreadsheet: "Event Name","Event Start Date","Event End Date","Event Start Time","Event End Time","event description or notes","URL applicable" "Meeting with John","2024-06-10","2024-06-10","09:00","10:00","Discuss project updates.","http://example.com" And here's the script ChatGPT made. It seems to get confused with things like Daylight Savings and/or Timezone but you can fix that manually. It took some debugging and I had to copy/paste my Mac's manpage for DATE into ChatGPT, but it got it eventually. The only...
Read more

Some rough notes on FreedomGPT and PrivateGPT

 The following are rough notes. I'll tidy them up when I get the steps 100% right. https://www.freedomgpt.com/ https://huggingface.co/mosaicml/mpt-30b https://files.mistral-7b-v0-1.mistral.ai/mistral-7B-v0.1.tar use case: wordpress viruses; recognising one https://www.icloud.com/shortcuts/8bcc2979d6e34a3d90a48c3f77f96d68 openai - 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 hugg tok - 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 https://developer.nvidia.com/blog/nvidia-releases-open-source-gpu-kernel-modules/ https://github.com/eosphoros-ai/DB-GPT privateGPT fix: sudo apt-get install python3-dev python3.10 -m pip install --upgrade pip sudo apt install build-essential python3.10 -m pip install hnswlib python3.10 -m pip install chromadb then python3.10 -m pip install chroma-migrate chroma-migrate python3.10 -m pip install -r requirements.txt pip install pycc How to download the models: Visit the Code Llama repository in GitHub and follow the instructions in the README to run the down...
Read more