5 ms·

> What am I missing? I'm going to guess (assume) you probably haven't worked in a 'real' business. A place where elbow grease still does the majority of the wo

> What am I missing? I'm going to guess (assume) you probably haven't worked in a 'real' business. A place where elbow grease still does the majority of the work and where Windows 8 was only just phased out. The killer app (to me) in case of GPT is GPT itself, not ChatGPT. ChatGPT just allows me to easily test use cases for GPT. There are many interesting use cases for those elbow grease businesses for GPT. For example: Data entry. There's still a lot of data entry being done from unstructured text. Where specifics like names and addresses need to be extracted from letters and emails and contracts and other stuff. I've worked on these challenges before using different strategies and GPT blows my mind with what it can do just by asking to grab this data and format it as JSON. Is it 100% correct? Nope. You still need people to review the data (depending on the use case), but that already saves tons of work. Categorization. Some companies still get tons of emails that need to be forwarded to specific departments. This is another thing that GPT does surprisingly well out of the box. And that's just GPT. There are many other legacy business processes that can be automated (partially) by other models that are coming out right now. Even just 'segment anything' that Meta just released is incredibly useful for many use cases I've seen in my daily work. Killer apps are always a combination of a technology to solve a real world problem. If you don't venture into the real world and only stay part of the tech world, seeing the killer app is very difficult and ends up leading to Juicero-like products.

naavis 4y ago
> There's still a lot of data entry being done from unstructured text. Where specifics like names and addresses need to be extracted from letters and emails and contracts and other stuff. I've worked on these challenges before using different strategies and GPT blows my mind with what it can do just by asking to grab this data and format it as JSON. Is this data confidential or something you are willing to send to anyone? If the former, you probably shouldn't be sending it to an AI company that retains the data for its own purposes.
- donkeyd 4y ago
  This is very confidential data, which is why the current implementation is run on-premise and of course I'm definitely not using production data to test GPT capabilities.
  - naavis 4y ago
    That's good! There have recently been news of companies leaking trade secrets and confidential data by sharing it with ChatGPT: https://www.techradar.com/news/samsung-workers-leaked-company-secrets-by-using-chatgpt https://www.techradar.com/news/samsung-workers-leaked-compan...
    - donkeyd 4y ago
      Well, in the organization I was doing this, they had similar issues before I started, with external consultants doing stupid stuff. Luckily, my team wasn't as stupid as these people and integrity was also very high. Which was very, very necessary considering the data we were working with.
version_five 4y ago
Right, questioning the value of a chatbot is a good indication someone hasn't had real job. Totally legit characterization that doesn't make the whole thing sound like a confidence game.
- epigramx 4y ago
  they mean they didn't have a job where neither ai or bad workers should be used.
- greeny7373 4y ago
  Real Job vs. outdated job. You don't want to know how many companies still get paper bills scan them and add them to their system half manually. And normal people without scripting experience never had the tools to just do a little bit of text analysis without tools like chargpt.
- donkeyd 4y ago
  They weren't questioning the value if the chat bot, they were questioning the value of the models based on only the high profile use cases such as a now (in)famous chat bot. That, to me, showed that they have a narrow view of the problems that this stuff could solve when applied outside of those high profile use cases. So I made an (explicit) assumption that they mostly worked in tech and not outside of tech, which limits the view of the world outside of tech.
- rlt 4y ago
  > Please respond to the strongest plausible interpretation of what someone says, not a weaker one that's easier to criticize. Assume good faith. https://news.ycombinator.com/newsguidelines.html https://news.ycombinator.com/newsguidelines.html > questioning the value of a chatbot Original commenter was questioning the value of the entire field. > hasn’t had a real job” This is different than “haven't worked in a 'real' business”.
  - version_five 4y ago
    Ironic given yours and others emphasis on my choosing "chatbot" to summarise llms as a reason to dismiss my comment, along with the rest of the pedantry. The upstream post dismissed / insulted the person for questioning value, which was what I called out. If you really had wanted to get into the "HN rules" game, you could at least have cited "don't be snarky"
stepbeek 4y ago
Data entry is a really interesting one to me. We’ve been replacing a legacy (read > 1 MLoC with no tests), system on and off for a few years. The original system had a ton of double entry or manual data entry and the human error rate is noticeable. If GPT could have automated this with a similar or reduced error rate then we would have considered it a win. The real win long term remains killing off manual entry any time it’s possible, but GPT offers a nice patch.
throwaway2037 4y ago
Nice call about "Categorization". I would put a lot of "Data Entry" in the same category. Excellent ideas!

naavis 4y ago

> There's still a lot of data entry being done from unstructured text. Where specifics like names and addresses need to be extracted from letters and emails and contracts and other stuff. I've worked on these challenges before using different strategies and GPT blows my mind with what it can do just by asking to grab this data and format it as JSON. Is this data confidential or something you are willing to send to anyone? If the former, you probably shouldn't be sending it to an AI company that retains the data for its own purposes.

donkeyd 4y ago
This is very confidential data, which is why the current implementation is run on-premise and of course I'm definitely not using production data to test GPT capabilities.
- naavis 4y ago
  That's good! There have recently been news of companies leaking trade secrets and confidential data by sharing it with ChatGPT: https://www.techradar.com/news/samsung-workers-leaked-company-secrets-by-using-chatgpt https://www.techradar.com/news/samsung-workers-leaked-compan...
  - donkeyd 4y ago
    Well, in the organization I was doing this, they had similar issues before I started, with external consultants doing stupid stuff. Luckily, my team wasn't as stupid as these people and integrity was also very high. Which was very, very necessary considering the data we were working with.

version_five 4y ago

Right, questioning the value of a chatbot is a good indication someone hasn't had real job. Totally legit characterization that doesn't make the whole thing sound like a confidence game.

epigramx 4y ago
they mean they didn't have a job where neither ai or bad workers should be used.
greeny7373 4y ago
Real Job vs. outdated job. You don't want to know how many companies still get paper bills scan them and add them to their system half manually. And normal people without scripting experience never had the tools to just do a little bit of text analysis without tools like chargpt.
donkeyd 4y ago
They weren't questioning the value if the chat bot, they were questioning the value of the models based on only the high profile use cases such as a now (in)famous chat bot. That, to me, showed that they have a narrow view of the problems that this stuff could solve when applied outside of those high profile use cases. So I made an (explicit) assumption that they mostly worked in tech and not outside of tech, which limits the view of the world outside of tech.
rlt 4y ago
> Please respond to the strongest plausible interpretation of what someone says, not a weaker one that's easier to criticize. Assume good faith. https://news.ycombinator.com/newsguidelines.html https://news.ycombinator.com/newsguidelines.html > questioning the value of a chatbot Original commenter was questioning the value of the entire field. > hasn’t had a real job” This is different than “haven't worked in a 'real' business”.
- version_five 4y ago
  Ironic given yours and others emphasis on my choosing "chatbot" to summarise llms as a reason to dismiss my comment, along with the rest of the pedantry. The upstream post dismissed / insulted the person for questioning value, which was what I called out. If you really had wanted to get into the "HN rules" game, you could at least have cited "don't be snarky"

stepbeek 4y ago

Data entry is a really interesting one to me. We’ve been replacing a legacy (read > 1 MLoC with no tests), system on and off for a few years. The original system had a ton of double entry or manual data entry and the human error rate is noticeable. If GPT could have automated this with a similar or reduced error rate then we would have considered it a win. The real win long term remains killing off manual entry any time it’s possible, but GPT offers a nice patch.

throwaway2037 4y ago

Nice call about "Categorization". I would put a lot of "Data Entry" in the same category. Excellent ideas!