HN user

BrianOnHN

1,181 karma
Posts18
Comments471
View on HN
news.ycombinator.com 3y ago

Ask HN: Are you interested in transitioning from SE to AI Practitioner?

BrianOnHN
1pts0
wfhbrian.com 3y ago

Obsidian Chat with your notes using GPT-4 and Embeddings

BrianOnHN
4pts0
wfhbrian.com 3y ago

Obsidian's New Smart Chat for Dynamic, Contextual Conversations with Your Notes

BrianOnHN
2pts0
wfhbrian.com 3y ago

GPT-3 and a Name Can Accurately Represent Your Personality

BrianOnHN
2pts0
wfhbrian.com 3y ago

Learn to write prompts instead of using Jasper

BrianOnHN
4pts0
wfhbrian.com 3y ago

Why Prompt Engineering Isn’t Going Anywhere

BrianOnHN
2pts0
wfhbrian.com 3y ago

The rise of working remotely and the need for skill guilds

BrianOnHN
44pts27
wfhbrian.com 3y ago

I Use Obsidian to Manage My Knowledge

BrianOnHN
2pts2
wfhbrian.com 3y ago

Perspective-Context-Action Prompt Writing Method for GPT-3 (Zero-Shot)

BrianOnHN
1pts0
wfhbrian.com 3y ago

What are Dynamic templates in Obsidian and why you should use them

BrianOnHN
1pts0
wfhbrian.com 3y ago

I use GPT-3 with Obsidian to explore prompt engineering

BrianOnHN
3pts0
wfhbrian.com 3y ago

How the Context-Subject Framework Can Help You Organize Your Notes

BrianOnHN
3pts0
wfhbrian.com 3y ago

GPT-3 at your fingertips with Obsidian

BrianOnHN
1pts0
news.ycombinator.com 4y ago

Ask HN: Why are you here?

BrianOnHN
80pts145
www.npr.org 4y ago

University of Florida bars professors from testifying in a voting rights case

BrianOnHN
379pts450
twitter.com 4y ago

Have something that NBC News' tech investigations team should be digging into?

BrianOnHN
10pts2
twitter.com 4y ago

Cyber Attack, Possible Serious Breach, at U.S. State Department

BrianOnHN
156pts133
news.ycombinator.com 5y ago

Ask HN: Does HN still effectively serve Y Combinator?

BrianOnHN
5pts2
ChatGPT Plugins 3 years ago

I've seen tweets that suggest they are using it, particularly, the prerelease GPT-4 was used internally.

I wonder if they fixed replying to Gmail threads. Last time I tried, without the NLP layer, it failed to do anything but create new threads.

To make our meetings more effective and efficient, we have created Mochary Method software. If you find the software effective, feel free to use it with your reports as well. [1]

Does anyone have any references for this software?

Typical reactions to the software are:

Meeting 1: “Why are we using this software? Let’s just use a Google Doc, it’s easier.”

Meeting 2: “Oh, I see why this software is better than a Google Doc. Let’s keep using it.”

Meeting 3: “This software is awesome. Can I please use this throughout my entire company.”

Seems interesting.

Edit: I think it's "CompanyOS" [2]. The homepage is nearly a replica of the Mochary Method website.

But, I haven't seen anything about how the software works. Does anyone have any information about it's features?

From the hiring page [3]:

Each is paying $150,000/yr for access to our product.

[1] "Mochary Coaching Methodology (CEO 1-1)" page https://docs.google.com/document/d/17AfqFdrx0lb6aYb786lY3a-1...

[2] CompanyOS https://companyos.app/

[3] https://mocharymethod.notion.site/Work-at-CompanyOS-419da897...

It's like the connection between objects in real life and objects in programming

What if this a more accurate representation:

"It's like the connection between dictionaries in real life and dictionaries in programming."

The neuron could be implementing NN theory in a way that is optimized for it's environment.

What about...

"We need a forum with accountability from people we work with and see regularly"

I know employers are already doing this.

Alternatively,

"We need a forum where locally connected people share the same avatar"

An additional layer of accountability being the theme in both.

It's basically blog spam and/or a social experiment to see how many automated articles they can farm karma with.

Author here. This is a false assumption. I'm only posting content I deeply care about, so the quality matters to me. Nothing is automated at all, it's an iterative process, all done by hand. And lol at "farm karma." Like, why?

I will also be writing more about how I use GPT-3 to help me write, so I wouldn't discount my content because I use GPT-3 just yet.

Yes, this was a "test the waters" type of post, and I will be elaborating on my ideas.

If we're at the inflection point of changes in how we work, which I think the explosion of remote work may suggest, then we have to completely reconsider what we "know" about work.

So, to answer your question, "how?"

We need to question every current assumption about work. The system is currently in a malleable state, but that won't last. More questioning assumptions, means more experiments, means more developments.

I have my hypotheses, and I'm working on those experiments. I hope to bring together more people that are interested in the same.

That sounds like an interesting book, I will have to check it out.

I didn't mean to suggest that a single skill guild is sufficient for any single person.

I believe the answer is somewhere in the middle of where we are now, groupings like subreddits, and what you're talking about.

Something where participation takes real effort, but in a way that complements other guilds such that members can "cross-pollinate" and grow as an individual opposed to growing into a mold.

AI-assisted, but not "generated" imo.

This topic is something I will write many more blog posts about. I have lots of notes on various topics, but they aren't in a publishable state. If I were in a better financial position, I would hire a ghost writer to help me work out the drafts. But, I'm not, so GPT-3 helps me with that.

I like this notion.

However, some sort of "fair & balanced" law would have to enforce this.

Edit: and to respond to sibling comment about fact-checking being heckled...

The mechanism here would have to somehow force a similar amount of views. For example, if a lie gets 1MM views, then the proof of the lie should have to gain 1MM views before the original author can gain leverage of the algorithm again.

Of course the new system will eventually be abused, however, it's a step in the right direction. And when that eventually fails to be recognizable, another set of checks and balances must be layered on top.

Under what argument should it be reasonable

To narrow the gap luck currently plays in life outcomes.

Currently, you are either born lucky, trust funds and generational real estate, or not.

More chances to be "lucky" would enable more opportunities for more people. And the OP suggests a negligible cost to make that happen.

Collecting such individual data via blood & urine samples (possibly fat biopsies & breast milk as well) on a yearly basis should really be part of a standard medical checkup procedure

No so long as sickness is so profitable. The incentives aren't there like you would think.

The other day I was think, it's actually the platforms where most of these communities that benefit directly from their existence.

And we know those platforms are tools for manipulation.

How much of the desire for these communities might be manufactured in the first place?