David Hariri

Beauty is not subjective. Applier of AI. Co-founder, chief architect @ada_cx. Prev. made @volleyworks and worked at @teehanlax.

[David Hariri @davidhariri]

LinkedIn is now just AI congratulating AI for AI-ing 😂

[Evan McCann @evansammccann]

@davidhariri it continues to get worse

[David Hariri @davidhariri]

Just checked my LinkedIn. At this point, why don’t they just post the ai slop for you?

[David Hariri @davidhariri]

.@cohere 1.0 UI had this aesthetic. It was ahead of its time.

[John Cosgrove @cameracoz]

@laurenloprete ngl been yearning for the windows 95 UI to make a comeback recently

[David Hariri @davidhariri]

High taste (in sw) is just lossless encoding of respect for a users time

[David Hariri @davidhariri]

@MikkoH Respect

[Eric Lombardi (EricForOLP.ca) 🇨🇦🚀 @EricDLombardi]

Can I just point out how funny this is, given the paper is owned by Canada’s wealthiest billionaire family, all of whom inherited their wealth.

Should we hate them for that? Of course not, because that would be gross.

But, it reflects our tall-poppy syndrome at its finest.

Personally, I’m glad that the Ontario Teachers Pension Plan is getting a nice payout, as will many others who took a big risk investing in SpaceX.

I wish we had more audacious entrepreneurs making big bets and generating great jobs and wealth here in Ontario. I want to create an environment that makes it even more likely.

The world is not zero sum. The size of the pie is not fixed. If we make more, everyone can get a big piece.

[David Hariri @davidhariri]

@EricDLombardi Welp, you just got my vote, Eric.

[David Hariri @davidhariri]

@ClementDelangue @ArtificialAnlys all* (not both)

[David Hariri @davidhariri]

@ClementDelangue @ArtificialAnlys Hm. Spiritually I want to agree, but if you can measure both by tok/s, $/1M input/output and the input and output is functionally equivalent, isn't it just up to the reader to make a latency:cost:intelligence decision?

[David Hariri @davidhariri]

"We used to claim outcomes".
https://t.co/0Ier5iVjVq

[David Hariri @davidhariri]

Recommended reading. Besides being a worthy profile of two excellent people (@bwertz and @angelatytran), it's also an unusually thorough recount of the early days of several startups and how important early stage investors are in actively supporting founders (far beyond the cheque).

[Evan McCann @evansammccann]

Read it here
https://t.co/iZ3hiId9vH

[David Hariri @davidhariri]

@ddebow Truly

[David Hariri @davidhariri]

@gozmike I’ll try again tonight. Mine stopped. No explanation. Hermes falls over when it gets 429s. Openclaw is an even bigger mess.

[David Hariri @davidhariri]

Here's your daily cup of reality

[Mitchell Hashimoto @mitchellh]

Fable is a good model. As with all new models, it is simultaneously excellent and entirely unremarkable (relative to other models). It is slow and expensive, and the "loops are all you need" discourse they are pushing is obvious in the context of someone using Fable-class models

What I've found so far is that for broad scope design (code architecture) tasks, Fable is unremarkable. Or, not better enough to justify its cost and speed.

But in highly targeted goal-oriented loops, it is another beast entirely. It is very slow but produces very good results.

I let it churn on optimizing a SwiftUI-layout resolver in Go I wrote and it was able to bring it down to an order of magnitude I could not reach myself (micro => nanosecond scale). But it took 2 hours and $40 to do it and I had to claw back some changes it overfit to Apple Silicon. Still, very worth it.

In comparison, for "implement this feature/change" iterative work, I ran head-to-head Fable vs GPT5.5 vs. GLM-5.1. They all produced equally acceptable final results, but GPT5/GLM did it in a couple minutes and Fable was churning away for 40 minutes. And GLM cost me less than a dollar, GPT5.5 ~$1.50, and Fable cost $9.

You can see that in this context, interactively working with an agent is nonsense. Its too slow. You need to write loops to keep the agent working and you probably want to highly parallelize the work being done. As with all things, I think a balance makes sense...

My sense is that I'd reserve Fable for targeted, surgical analysis and work. Not for daily driving everyday tasks.

I'm going to keep spending a shitload of money (relatively) and maining Fable for the rest of the week to continue to judge, will report if anything changes. I'll continue to head-to-head as well.

[David Hariri @davidhariri]

@gozmike Which harness? CC?

[David Hariri @davidhariri]

@skanwar @borisc @torontolibrary @aratisharma @TPL_Foundation Same! We go to the Pape branch every weekend.

[David Hariri @davidhariri]

@skanwar @torontolibrary @aratisharma @TPL_Foundation ✅