jaredsohn

Also for acquisitions, often you'll have to work at the acquiring company for some time to get money from your options. Or might get options in the acquiring company instead (which again are worth nothing until some future possible equity event which hopefully translates into cash).

Some use cases for figuring out what is possible are when you're building something that relies on APIs (such as a chrome extension and you're not familiar with how they work / what they do or just what permissions are needed) or requires that you have certain data specific to each customer.

is backwards even

Lots of useful information.

I guess you need to practice what this suggests rather than just sharing it with someone, though. If you send someone a link of 'How to work with "stupid" people' they might think you're calling them stupid instead of trying to improve communication.

It would become super amusing if you ask Grok how to do something like this and the recommendation becomes 'Ask ChatGPT'. :) But I'm sure people working on training datasets are filtering out such responses.

I think the right answer is that people who care about these things should use a client that rewrites queries via an LLM and clearly show the changes it makes (not sure if this exists yet but I'm sure it will). This also would handle a few other problems I've run into such as having multiple WHERE clauses or doing a WHERE prior to a JOIN and it could also allow using plain English to describe what you want to query.

I've found LLMs do a really good job of writing/cleaning up SQL.

I think you need to create a project file and then it stores things there.

For a long time I would get paranoid about accepting Mac updates which would require a reboot because then I'd lose my undo history and then I discovered that this is all I would need to do.

https://hnreplies.com/ is an option (although using it means tying an email to your Hacker News username within their database which some people won't want.)

You can also just look at the threads page occasionally.

Disagree. AI can increase productivity a bunch of ways in this scenario:

1. "Help me understand what this code does." Both via text and by building diagrams.

2. "Help me understand how to solve this problem or how this tech works."

3. Normal code-completion type things

4. Helping build scripts that you can use to apply changes.

Also, in general you can ask the AI to maintain the existing style.

I think AI can help reduce the burnout. Send it some code that you think may be relevant and ask about it and then verify after you see the AI's explanation. Of course AI can burn you out, too, if it doesn't find something useful but it can be a time-saver for this purpose sometimes.

Basically try to take advantage of when verification is easier than what you're asking AI to do for you.

If I google the phrase now, it includes Biden.

>why won't Google/Mozilla prevent this by making an extension and a person's account inseparable

This can be gotten around easily by making a separate Google account for the extension. It would require using gmail rather GSuite (without transferring over the entire GSuite domain.)

Surprised to see no mention of how this impacts people who purposefully write scam messages poorly to filter out people are less likely to get conned. Curious if this starts making traditional scammers less efficient.

>Are you diligent enough to remove your sensitive logging/printf statements EVERY time

Yes, when a linter is set up that fails when printf debugging is found

A similar use case for me - I wrote some technical documentation for our wiki about a somewhat complicated relationship between ids in some database tables. I copied my text explanation into an LLM and asked it to make a diagram and it did so. Took very little time from me and it was fast/easy to verify that the quality was good.

OpenAI has already built some of this but maybe it requires a paid account.

https://mashable.com/article/chatgpt-chat-history-search-int...

It also has projects (says is for Plus, Team, and Pro users). https://help.openai.com/en/articles/10169521-using-projects-...

I've found processes like this to work better, too. Basically, the one big pr is like building a prototype to throw away. And the benefit is it has to get thrown away because the PR will never pass review.

OpenAI acquisition/synergizing/rebrand to zombo.com incoming and then we'll complain about them ruining Zombo.

FaxGPT as well

Yeah, you're right. I did a little investigation before posting that and it seemed to handle some movement within the file but after looking at it more closely I see it is not fully robust (even within the same file it misses some earlier changes).

This page provides some other techniques: https://tekin.co.uk/2020/11/patterns-for-searching-git-revis... but I think they have their own limitations.

Here is how you can do this via the commandline for one or more lines:

git log -L 40,50:app/models/user.rb

I think SELECTs without WHERE clauses are fine

Maybe 'public AI' is a better term than 'open' ai

I'm looking forward to developers setting up LLM prompts to make their code seem more complex and like it required more effort.

If you build something that doesn't solve problems with impact to the business, your real productivity is zero. How does this account for that?

https://blog.pragmaticengineer.com/the-product-minded-engine...

When I last looked it seemed like cloud costs were much higher.

How expensive is it to run these in the cloud? Last time I looked into this it looked like you have to pay significantly in server costs to be able to run a local LLM.

Presumably someone will notice the bug in both of these scenarios at some point and it will no longer be treated as safe.

Good tests reduce your cognitive load; you can have more confidence that code will work and spend less time worrying that someone will break it.

BTW, I think above are the best metrics to use for tests. Actually measuring it can be hard, but I think keeping track of when functionality doesn't work and people break your code is a good start.

And I think all of this should be measured in terms of doing the right thing business logic-wise and weighing importance of what needs testing based on the business value of when things don't work.

Hate to tell you but there is nothing stopping people using AI from doing the same thing.