lazharichir

> Put yourself in the user's shoes. On what basis should one pick a model for a given task?

Create a realistic eval suite for your use case using your actual data that you manually label, with easier, more complex, rarer samples, and run each model you are considering against it and decide based on the usual confusion matrix, cost, latency, or whatever metrics matter to you. There's really no other way to know – and you might be surprised that cheaper models can often perform your task(s) relatively well at a fraction of the cost.

I am using A2A at work. It's a bit like the "microservices architecture" for agents... Allows you or teams to develop agents independently and have them interact as and when they need to. No major hurdles so far.

What can you NOT run on this, it's not very clear? Is it like MicroVMs on steroids where you can run more binaries than the strict minimum?

Definitely the feature but I'm sure Gemini is seconds away (figuratively) from invading Chrome and if it has an agent mode itself, it will eat everybody's lunch in the browser space.

Just excited about Agent Mode, I hope it delivers.

Quite happy when I got to play with it – but I do prefer Inngest as the code is hosted where our actual app is hosted; not on their server. So it's just easier (but it's a tradeoff).

Good tool, good tooling, congrats to the team!

From my (limited) understanding you cannot really switch these off inadvertently as they require a couple of actions in order to be switched off. So it would mean one of the pilots switched these off (and they were a few seconds later switched on again but it was too late).

But there was audio, too, and one pilot asked the other "why did you switch these off" and the second one said "I didn't".

Was there are third one in the jump seat?

You can also do gemini flash lite for a subset and then batch the rest with flash or pro

Bedtimely[0], a personalised bedtime story generator for parents of kids aged 0-10yo. Scratching my own itch with it, especially with the participative interactive stories where the parent and child decide what happens next.

Work in progress...

[0] https://bedtimely.com/

That's basically how I use it though I cannot wait for Gemini to be a click away in my browser with the current screen/page(s)/tab(s) embedded so I can ask it stuff about the current long article/documentation page. We're becoming very, very lazy.

to be fair, i'd take tame over horrendous and unparseable screen any day.

Like taking on a gazillion of contract work?

Building HatchAStory for my own needs as a busy dad — to generate fun, personalised bedtime stories in seconds.

https://hatchastory.com/

Built a backend and web version but now focusing on an Expo/React Native app (my first ever).

I assume they mean that you send an Array<Message> to the model instead of a single system and user message.

Good small fun project

The problem is user experience. It's still very much a chatbot... To justify that amount, it needs to integrate a lot more with an employee's day-to-day tools such as code editor for SWE, the browser for quickbooks, words/sheets/powerpoint, salesforce, HR tools, and so on.

Still the best product/experience out there across the publicly available products. Bard is no GPT4, still, to this date.

I think it's also the surprise of how sudden it unfolded before the public's eyes. And that, seemingly in an organisation that's figuratively on top of the world currently, and pushing through.

> Most coworking places are clean, organized, have snacks, do not have Wi-fi issues

You would think so, but this is incorrect. Every single WeWork you go to, across the globe, offers the same fantastic experience from "feel good vibes" to functional meeting rooms, wifi, and ad-hoc services.

But even if there are local gems, it becomes another chore to browse listings, read reviews, spend time deciding between Coworking Space A and B and C. With WeWork and Regus/Spaces, you save a lot of time and can just get there, and get busy.

> The customer you describe "picky and willing to pay double normal rates when travelling to foreign countries" is quite of a niche use case.

Not OP but just like luxury hotels are niche, WeWork is not your local coffeeshop and does cater to a different crowd. Whether that's profitable for them or not, looks like not for now.

Not true in my opinion. It's mostly that what is perceived as success may be birthed out of sacrifices, compromises, and loneliness. Many bootstrappers I know have managed to be successful but lost loved ones whilst getting there.

It does. It have to be this way, but it often takes obsessive behaviour for many to get results. And that's just unhealthy, whether you succeed or fail.

Very curious on how this versatile project will turn out. Google, with all of its collaboration tools like Sheets and Docs, really sucks at "Teams" (task and project management, for instance).

Probably because it's hard to fit all the use cases G Suite customers may have (tiny teams, giant enterprise, growing startups, etc).

Yet, I've always wanted that with my small team working off G Suite mostly. Always having to use an external Project management is fine, but having it done from Google itself would be a great touch. A little like Microsoft Teams/Sharepoint.

I hope that Table can bring us a little step closer.

Erm, the explore tab in Sheets is pretty awesome for small businesses and people unfamiliar with Excel-like formula.

Not sure when that was released, though!

It really isn't "done" if you tell people they probably should look at alternatives instead. Sounds more like a soft deprecation to me.

But regardless of semantics, they're being transparent on the project's status.

1. When I learnt TypeScript after having used mostly PHP and JS until then – not sure how I lived without stronger typing

2. When I started reading programming and software engineering books after having just "done it" – this brought so much thought and structure to what I used to improvise

Do you have examples of what you mean by red tape from the EU that prevents a bootstrapped startup from thriving?

Same. EE 4G in CENTRAL LONDON with a SIM-card Home Router and the speeds are horrific — mostly 5Mb, and maybe 15Mb for an hour or two each day.

Used to be consistently 50Mb+ back in the days.

AWS Status page[1] shows a couple of issues in eu-west-2.

Amazon Elastic Compute Cloud (London)

Instance Connectivity

3:21 AM PDT We are investigating instance connectivity issues in a single Availability Zone (euw2-az2) in the EU-WEST-2 Region.

Amazon Relational Database Service (London)

Small number of instances unavailable in a single Availability Zone

3:36 AM PDT We are investigating connectivity issues affecting some instances in a single Availability Zone (euw2-az2) in the EU-WEST-2 Region.

[1] https://status.aws.amazon.com/

I am on GCP but I usually request/allocate more and once in production for a few weeks, I can turn some things down and others up. That is if your product/setup is new and you aren't sure how everything will interact once live (in terms of perf).

Otherwise, for Lambda there is on Github a repo with a script that will publish your code in every configuration possible, run stress tests, and only keep the most optimum (perf/cost).