dzonga

I have been considering using my other laptop as a place to run my agents in production since 1. my laptop is more powerful than the vps I can rent for cheap.

then just use Cloudflare tunnels to connect to the laptop.

for the downside - countless cultures have had preventative mechanism for that.

if your biological child is an idiot/wasteman - then you could adopt - the Roman Empire ran on adopted children, with biological children seen as wasteman.

the other way it's done e.g with Japanese businesses - again adopt your son in law and give him your surname.

the thing about giving helm to the sons - is usually you would've had a whole lifetime to evaluate the candidate 20+ years. do mistakes happen yeah, but honestly the risk is overblown.

for a lot of teams that do pair programming etc and pairing with agents - this makes sense.

hell dare I say better than the pull-request model - since you don't have to wait on people to review the code and deal with the nonsense of merge-trains.

there's nothing wrong with leaving things to your kids to run.

that's the point of the effort after all.

warren as smart as he is was tryna go against the grain of thousands of human evolution. if you train your kids - then most likely they would've values similar to yours. hit or miss but most times a hit.

open-ai have proven they can make good / decent models but business strategy is just spray and pray.

they need to pick a lane and optimize for it. coz at their size they can't serve the application layer (a.i startups who can fine-tune models will eat their lunch)

if they gonna do a consumer play - then go ham on that.

otherwise they're gonna get caught in the dreaded middle valley.

very few people comprehend - how much of an asteroid level event for western AI labs this is.

china has cheap abundant power, now they can make their own inference chips (which was supposed to be a chokepoint), their models yeah can be 6 months behind the frontier - but most people don't need frontier models - small models r more than enough.

my only wish was labs like Mistral would make their own inference chips or partner up eg with established / new chip makers or companies like Oxide.

the open burial started when zAI served their latest model on all Chinese chips.

now we r just noticing the grave getting dug deeper.

black swan events.

this is one of one of those things - were in the current era either a cloud provider should provide automatic backups in another geographic zone.

if you're in us-east, then your back-ups should ideally be in eu-west + africa for redundancy.

yeah the performance, reliability & ecosystem of the JVM is unmatched.

if you gonna work in a big team or need a project with lots of devs then yeah go for the JVM.

but if you're doing things on the smaller / small scale side. - just use JS/TS or python. you benefit from cheap runtimes such as Cloudflare workers.

people on the ground know that small models are enough, since llms are good at directed work (i.e handholding) not the let loose go wild that the labs try to hype on.

the only thing that few people are willing to admit is that humans are the bottleneck as humans are needed to handhold / verify output - which puts a dent or might I say pause on the excessive valuations of a.i companies as that's against the narrative.

AI assisted coding all the way.

for someone like me who doesn't like to write tests - I could easily have an agent write a test for me.

I could lay the bricks then also have the agent do ornamental design that I'm not good at.

the tragedy is that architects (experts) prefer the modernist buildings. while common people prefer the classical buildings which makes cities more livable. the other tragedy is before 'regulations' normies designed buildings - they didn't have architecture degrees. don't know if that's still possible anymore.

& buildings shape lives.

the only way to reverse the modernist trend is to have laws that incentivize classical buildings otherwise we gonna keep getting ugly buildings.

if she was eating sweet potatoes how can that be ? unless she wasn't eating the sweet potatoes at all ?

I have a shitter account, I don't use the app but just the web app on safari iOS.

recently with videos I can't change playback speed. so I have to use xCancel.

before I just used to use Tweetdeck. but that went away or went pro.

oracle could've cruised nicely but nope greed and Larry Ellison's ego (add his son's too) - they bit off more than they can chew.

this is a welcome ruling.

most of these search engine like companies including metasearch etc benefit from scrapping data on the onset, but then want to bring up TOS when they get scrapped.

in the age of agents - if your agent does work on your behalf on a particular site that should be legal.

btw does restate run on Cloudflare workers ?

in the age of agents, surely durable execution is a must.

but the money being raised - what is it for ? go to market ? sponsorship & I like that temporal sponsored Crystal Palace.

we've so many good durable execution engines.

this is the type of VC pollution that will cause companies to end up being acquired by an Italian silverware company (bending spoons).

or underestimating the value of partnerships.

nike thought they were the storefront. reality and history proved otherwise.

I'm not saying the bet is against Nvidia the company as in how it executes, or enters new markets.

the bet is against the high margins in inference - when other capable players have entered the market as models become commoditized and the inference serving chips as well. Cerebrus ai etc are already showing custom silicon can make a dent while being served cheaper & faster.

this is why people are going native / retail media. & some people i.e those with a lot of money are going to CTV as well.

search ads have negative ROI.

Facebook/Instagram & linkedIn work to a certain extent.

80-90% of AI costs are related to inference. Once zAI served their new model entirely on Chinese chips for free then the nVidia jig is up.

who is going to pay back the money nVidia invests in AI labs, AI datacenter companies if the models are being served dirt cheap.

the Chinese are not the only competitor - Amazon with their Trainium, Google with their TPUs etc.

Nvidia might have a moat on training but on serving it's gonna be a blood bath.

But for now they're capturing 80% of all the AI spend so they gonna keep making money.

2.5 flash was also good to use with as the llm layer for voice products.

but google gonna google.

the Houthis won't desert as they're fighting an ideological war vs the Saudi backed forces which are just mercenaries.

idealogical fighters fight to the death. & in almost every case besides the nazi's etc they always win. whether the ideology is right or wrong that's a conversation for a different day.

what's worrying is they knew they were going after an operator whose organization they recognized as competent in terms of carrying out operations, recruiting and passing down information.

& they sat on it.

that's gross negligence honestly - if your adversary seems competent - you've to destroy them at the earliest chance before they amass more resources.

that's my only gripe with deepseek honestly

em-dash by Cloudflare looks interesting in terms of replacing Wordpress.

by garnering enough of an ecosystem to cause a migration will be the bigger challenge.

of course some people will say in the age of agents will people need to visit websites ;)

in short they're TRUST networks.

which is also why crypto bros get confused.

I agree. things like tailwind etc should come out as by-products of products making money.

react was not facebook's product. rails was not 37signal's product. Kubernetes was not google's product.

blaming AI is not facing the reality of how tech products / open source is financed in our industry.

for a first foldable iteration - apple has hit it out the park compared to other foldables on the market.

when the google & Samsung ones first came out they were plagued with problems.

but I think apple should've also had a foldable in the form of a razr ultra since they already have an iPhone air.