Haha <https://www.langpal.com.hk/subtitles> focuses on subtitle generation but just for Mandarin and Cantonese but the underlying pipeline is made in a way that doing english would be a appending an if or 2 or 3\
\
For me though I'm more focused on my niche so I can deliver the best experience for that niche!

Founder @ land.org.au & langpal 语朋

Devon

I definitely have quite a few thoughts:

- It's super cool to have an idea and see it to fruition, especially cool to see new hardware popping up
- But I think LLM-based hardware is in a VERY precarious situation due to all these issues surrounding Humane and Rabbit, so consumer expectations are much much higher
- Avi has good Twitter clout but I don't see the founder-product fit, he previously made/helped make 2 very popular and socially positive web apps but has (seemingly) no hardware experience (unlike Palmer Luckey of Oculus for example)
- I think the video has super high production value but I'm skeptical it works anywhere close to how it currently functions; it could be a look into the future, but due to the above consumer expectation, unsure if that's a good idea
- The pricing is too low for what you get; even if it functions just like ChatGPT, $99 for infinite inference to Whisper & a GPT3.5-4 level model seems wild
- Buying a domain for $1.8 on a payment plan is either playing 4D chess (if the product exceedingly delivers) or a bit silly if it flops; risk it for the biscuit ig haha

SWE @ Govt / Founder @ land.org.au & langpal 语朋

I’m actually quite a bit of a foodie but it conflicts with efficiency most of the time.

Prized meals:

Onigiri

When I first moved I had to optimise my budget and onigiri (w/ a rice cooker) is a super afk and cost effective method with super high versatility.

Tuna with Japanese (kewpie) mayo and even some egg based rice seasoning if you’re fancy 

I got too obsessed with them to the point where I probably ate 600 in a year. So I’ve gone cold turkey.

Cold soba, eggs & miso

This is my current everyday lunch strat.
Soba can be cooked in minutes if you have a kettle.
Miso is done in seconds with already boiled water.
For the eggs I have a strange egg boiling device from Kmart (amazing purchase)

To complete the meal I top it off with soba sauce and some water.

🤌🤌🤌🤌

SWE @ Govt / Founder @ Non Profit

@Alex my thoughts exactly, I did a demo with a few phones and the meta glasses, and even though it was a bit sus - this form of communication with AIs is the way forward!

I won't show the demo until I can trick it into calling my alt FB account "Chaat Geepeatee" though 😂

SWE @ Govt

The fail is my perpetual reasoning, 2am crunchyroll isn't optimal

Wins:

- Did a talk at Google Melbourne w/ like 60 ppl-ish
- Bought a sick cardigan at Uniqlo Sydney
- Getting support from that search company for my non profit

Fails:

- What's a gym?


- Sydney trip was a huge fail due to bad timing with many friends/business contacts

Taro’s biggest value proposition is being able to engage with a super active and passionate community.

There are certain topics that may not apply outside of the US and at times it’s tricky to balance live events but still being able to just ask any ￼question you want to know on here or in the slack is hugelyyy beneficial.

There’s a popular book called Outliers by Malcom Gladwell, which goes over the situations and events that led to hugely successful outliers becoming who they are.

It can be kind of depressing in a way but you can reverse engineer the stories of success to influence your own life.

And a HUGE factor of that is their network/location & social circle.

I’ve only lived in a small town in England and a small city/towns in Australia (until recently) and the lack of opportunity in these places is hugely impactful on each individual.

Not too long ago the concept of asking a Googler if they could check out my portfolio and that they’d say sure, would be completely unfeasible and not realistic.

Taro (and adplist.org) can enable that, taro more so as it’s a deeper social connection that’s established as opposed to a specific mentor-mentee relationship.

I generally have the opposite issue of having too many ideas and not enough time. 

Recently I’ve been open sourcing a lot of these projects and have an open sourced-open source Trello board.

Feel free to leverage: https://trello.com/b/6sFAveoP/dcrebbin-open-source

Wins:

- Released <https://upit.quest/> (open source AI leetcode coach)
- Eating consistently healthy

Fails:

- My old govt dept innovation team still hasn't improved despite my best efforts to help them outside of work
- Current day job work is hella boring 

What's...an...equity grant...\
\
No but seriously that's completely wild! 🚀🚀🚀

Google Hackathons are pretty cool: <https://developers.google.com/events>\
\
Local GDGs (Google Developer Groups): <https://gdg.community.dev/>\
\
Searching for Hackathons via <https://www.meetup.com/> \
\
Game dev hackathons via <https://itch.io/jams>

I can only comment from an Australian fed govt perspective:

Our federal government has lots of programs for both high school and university leavers where people can apply for these programmes and end up getting offers from a variety of different government agencies.

Depending on which a pathway you go through also depends on the level you start at.

These pathways may also not guarantee you an engineering role and so that is dependant on the pathway, your degree, your experience, the availability of roles etc (e.g: could join as a business analyst).

**Pros**:

- Incredibly stable
- Relatively good pay for that level of stability
- A lot of horizontal movement
- Low barrier of entry/not as much competition

**Cons:**

- The tech stacks are generally not great to work with and highly dependant on which b2b company had the best marketers to push said stack
- No high level IC role
- No stonks in your TC
- Higher average age so difficult to relate to your colleagues sometimes

Could go on for years but the opportunity changed my life for the better, it's really a question of what you want to get out of the role.

**Fails:**

- Failed a mock interview for meta to the point that simple addition was a nigh impossible task

- Showed up 10 minutes late to a meeting to meet my new manager whilst WFH due to stress based insomnia

**Wins**:

- Previous govt job's CIO responded back to me on an innovation I made whilst there that could improve upskilling for 25k staff and could lead to a positive impact 4 the aus public
- Google AI researcher working on Indigenous language research responded to back to me after I followed up (my first correspondence got lost!)
- Made new friends! 🤠
- Reached 150 github stars on a project!

As the organisation expands, and as each layer of hierarchy adds a level of abstraction, how do you maintain the core founding mission?