5x Founder Runs His Startup Solo With AI Agents

the best data filter may be **no filter**, with projections suggesting the crossover for internet-scale pools lands around **1e30 FLOPs**

The resulting model, SU-01, supports stable reasoning on difficult problems with trajectories exceeding 100K tokens

METR pays human programmers a minimum of $50 per hour, so getting a baseline for a single 160-hour task would cost at least $8,000.

At this rate, the phone in your pocket will run today's frontier models before you upgrade it.

I'm a little--maybe more than a little--concerned with trying to encapsulate every possible stylistic variation in every API into locale identifiers. A certain amount of this, sure. But this model won't scale and it will be difficult for developers and customers to use.

we've come up with micro designs and processes which when repeated should enable assemblies to grow exponentially to create critical campaigns that can take over councils and governments

It's going to be so fun to scale when you finally want to add more unsubscribe links for different types of sequences

Many of the neighbourhood organisations were able to support and initiate new projects and busi-nesses

Scalability – in the data/metadata stored;

Scalability

“multiple storage”

In the case of buffering, you could perhaps accumulate incoming messages in an array and periodically flush it on each requestAnimationFrame, rendering that list into the DOM. Alternatively, you could buffer them for some duration, like every second. If you’re appending them into an existing table, you’ll probably also want to use some form of table virtualization because rendering 100k rows is also going to be a huge bottleneck — and a form of backpressure!

Forwarding events from the native element through the wrapper element comes with a cost, so to avoid adding extra event handlers only a few are forwarded. For all elements except <br> and <hr>, on:focus, on:blur, on:keypress, and on:click are forwarded. For input and textarea, on:input and on:change are also forwarded. For audio and video, on:pause and on:play are also forwarded.

gRPC can use protocol buffers as both its Interface Definition Language (IDL) and as its underlying message interchange format

The gRPC infrastructure decodes incoming requests, executes service methods, and encodes service responses.

We instead scaled up our ElasticSearch cluster using Google Container Engine and with it’s powerful aggregations and facet processing our needs were easily met.

Scalability is the problem you want to have, and sooner rather than later, but maintainability is the problem you’re definitely going to have, sooner or later.

Popular architectures deal with heavy traffic loads by adding logic to cache popular views and resources.

Explicitly exposing any attributes that might get overridden by a parent seems impractical to me.

It would be tiresome - and bloated - to include a class pass-through for every component or assigning custom properties (from the RFC linked) for all potential properties on every component, just in case it's gonna be used in layouts that requires it. Wrapping them in a wrapper div is certainly an option, but potentially creates 100s or 1000s (long lists, several lists etc.) of new elements in the DOM slowing down low-end devices.

Your tooltip component will have to wrap your image with a span tag or something, it can’t just add events to its children. And if you are adding multiple actions to it you will have to wrap it multiple times.

<Concern1> <Concern2> </Concern2> </Concern1>

vs.

<img use:concern1 use:concern2>

You'll have to create a new component that brings in the functionality of both. TooltipButton, TooltipLink, Link, and TooltipRoutedLink. We're starting to get a lot of components to handle a bit of added functionality.

You can imagine the ensuing combinatorial explosion if we needed to add borders or box shadows or filters or what-have-you.

For the tooltip example, if you had a whole bunch of tooltips on different elements, it would be annoying to have different event listeners and "should it be shown" variables for each one.

You must: reference each element you are extending using refs or an id add code in your oncreate and ondestroy for each element you are extending, which could become quite a lot if you have a lot of elements needing extension (anchors, form inputs, etc.)

You can help ensure your RFC is reviewed in a timely manner by putting in the time to think through the various details discussed in the template. It doesn't scale to push the thinking onto a small number of core contributors.

It’s also been a year where we’ve had to acknowledge that we have reached the ideal of what we can achieve as a consulting firm in our current configuration, and — more importantly — that we don’t wish to transform the organization into something larger. As with most things that bring joy, the kind we experience within the confines our our tight-knit company are particularly problematic to scale — especially when faced with impending remote-working realities — so we’re opting to go the “love you and leave you” route instead.

At Google, our attitude has always been that users should be able to control the data they store in any of our products, and that means that they should be able to get their data out of any product. Period. There should be no additional monetary cost to do so, and perhaps most importantly, the amount of effort required to get the data out should be constant, regardless of the amount of data. Individually downloading a dozen photos is no big inconvenience, but what if a user had to download 5,000 photos, one at a time, to get them out of an application? That could take weeks of their time.