Comments by "TJ Marx" (@tjmarx) on "Fireship" channel.

  1. 82
  2. 37
  3. 27
  4. 12
  5. 10
  6. 7
  7. 7
  8. 7
  9. 7
  10. 6
  11. 6
  12. 4
  13. Largr generative language models aren't real AI. It's silly to treat them as if they are, they're just chat bots. 1. Will become increasingly less of a problem as things improve. Given it took us 50 years to improve chat bots to this point, how long it will take to improve them to be competent is unknown. It will eventually happen though, you just might be older than you think when it does. 2. This is actually a very silly point. Who gets blamed when third party SAAS messes something up? Why would you assume rented generative models would be any different? 3. You don't get more jobs out of a requirement for experience. Capitalism doesn't work that way however, it doesn't go "oh well we could just keep the status quo or we could use this trendy thing that slows down productivity and will increase costs, guess we'll use the latter". It's not going to be used in a commercial sense until it's good enough to reduce costs and increase productivity. Particularly when it's a paid service. The goal of capitalism is to reduce costs to as close to zero as possible (see slavery) whilst increasing price to consumer as high as the market will bare for maximum profits. If a tool doesn't help with those goals, it won't be used. But it's only a matter of time before it will help with those goals and that's what this video is talking about. 4. This point is puzzling, and I feel a little embarrassed for you having made it. It's not like everyone has stopped writing and publishing code, there will always be new code to scrape. It's like an organisation saying "well, I guess we already hired the cream of the crop programmers so all the new graduates should be ignored". But even if no new code was written, ever, if it already has all the good stuff, it already has all the good stuff. The model will improve. The training methodology will improve. As they do that dataset will become increasingly useful. Nothing is static, the point is not ChatGPT 4 - Turbo. The point is ChatGPT 25 - Ultra. It doesn't exist yet, but given Microsoft are all in and it's a useful tool in other areas it likely will become a reality in the future. Microsoft has an incentive to make it happen, it will reduce THEIR costs and increase THEIR productivity. 5. What about regulation does anyone imagine will prevent programmers from becoming obsolete, or near too eventually? Seriously, there's no proposal in any currently proposed or discussed legislation to prevent generative models from displacing workers. The legislation is centred more around giving I boundaries on what information it can give the general public, stopping it from disclosing secret or proprietary information, preventing crime, etc. The government doesn't care if programmers become obsolete, that's desirable because it increases the profits of the companies, thus it makes the economy more desirable. Personally, I don't think programmers will become completely obsolete. Generative models are just a tool, one that needs to be operated by an intelligent and creative mind. How, what this tool does is increase productivity to the point the workforce required to achieve a project outcome is dramatically smaller. Like, 1-10 team members. That makes future demand far lower than supply and will thusly impact salary. What matters is the timeline. If it's going to happen in the next 5 years, that's an existential problem to anyone studying CS or Software Development right now. But computers writing software has been a goal since the 70s. It seemed imminent in the 90s and then it wasn't. So for all we know it could be another 50 years away and meaningless to anyone in uni today. Nobody KNOWS the timeline. People can speculate. They can make informed educated guesses. But that's how the stock market works too, so I wouldn't put all my eggs in any basket. The reality is, no role is stable and it never has been. Some hypothetical new proprietary technology we haven't heard about because they were keeping it secret from competitors could emerge tomorrow and wipe out the need for teachers or funeral directors. That's always been the case and history is littered with examples and roles that no longer exist. The smart worker never puts all their eggs in one basket. They never become so specialised in just one thing that if that role were to disappear they wouldn't be hireable. The smart worker remains versatile enough in their skill set that should their role or even entire industry evaporate they can rapidly pivot to something else.
    4
  14. 4
  15. 3
  16. 3
  17. 3
  18. 3
  19. 3
  20. 2
  21. 2
  22. 2
  23. 2
  24. 2
  25. 2
  26. 1
  27. 1
  28. 1
  29. 1
  30. 1
  31. 1
  32. 1
  33. 1
  34. 1
  35. 1
  36. 1
  37. 1
  38. 1
  39. 1
  40. 1
  41. 1
  42. 1
  43. 1
  44. 1
  45. 1
  46. 1
  47. 1
  48. 1
  49. 1
  50. 1
  51. 1
  52. 1
  53. 1
  54. 1