- Glitch | Product Engineer | Ireland & Remote (EMEA preferred) | glitchads.ai
Glitch is building the AI agent for Advertising
Glitch let's Businesses create Ad Campaigns in 5 minutes (Down from 5 days) that perform 5 times better than traditional agencies. Our AI makes high performing ads available to everyone.
We need Engineers who want to tackle hard problems like treating the real world as context for LLMs, understanding markets in a few seconds, and dealing with the complex ML algorithms that underlie digital advertising.
We are a team of 8 and a mix of Startup Hustlers and ex-Google Ads Engineers. We've raised Seed and are doubling our customer base every quarter.
We love python, react, tailwind, OpenAI, Google Cloud, and Claude Code. The ideal candidate is running 10-15 instances of Claude Code at once.
Email me @ kingsley @ glitchads.ai
- Sep 1: generate requirements, generate design, generate task list
Step 2: run the tasks in claude code in parallel...
- Microsoft can afford to make a less profitable product for search than Google because if Google competes it’s a net win for Microsoft.
AI has many other profitable uses for Microsoft but specifically using it to compete with Google Search seems like a poison pill.
- It’s far more 5D chess than that.
Search costs nothing to run relative to ad revenue. Microsoft makes each query require 100x the cpu usage because users expect an LLM answer for results.
Microsoft’s share of search goes from 1% to 5%. Their cost go up but their sale of ads increases and they get valuable IP.
Google loses 5% share of market but its costs go up 100x.
Google can no longer finance its other bets like Cloud so effectively.
Microsoft meanwhile has a more compelling cloud offering.
Google starts to lose more ground on Cloud.
Amazon (not an AI company) lose ground to both.
Classic Art of War, if your opponent is strong, attack somewhere they are weak.
- It absolutely is. Googlers here will know that there was an internal version of ChatGPT that got canned because of halucinations.
- People pay for holes not drills.
- Strangely enough if you asked Europeans if they wanted high American Salaries they would also say yes.
The question is loaded and I guess 21% can think it through (or their company already gives them better).
- I've done two large search projects in the last year. One with Postgres search and one with elastic search. The elastic search version was trivial to configure, and was embarrassingly superior to the point where I wish I had never considered Postgres.
- Disruption typically happens where it's not only more efficient but cheaper too. The number of people you displace is going to be a fraction of what you make.
If you do end up making as much as the original industry it's because you are doing MORE than the original industry in which case you now have to question if the original industry didn't deserve to be destroyed.
Everything that can be software, will be software, what needs to happen is that the number of things that are software needs to outnumber the things that exist now. Software will create work for people to do as long as it keeps pushing what we can do as people. People need to adapt to a world with software at the heart, everything else is very dangerous.
- What: A better way to find github repos relevant to your interests. Works by compiling all the starred repos of contributors and . Currently in early alpha at https://github.com/KingsleyKelly/GitStar.
Why? I went from having about 10 starred repos to 300, and found loads of useful libraries and tools along the way.
- Ebola is actually too aggressive. It tends to kill and incapacitate people so quickly that they don't have time to spread it to others.
- Google+ has it's users, but to me it always seemed like the ultimate example of building something no one actually wanted.
- People are saying this is pedantry, but there's a good point being made here. Coding passion is a deep cold analytic burn that is nothing like what people assume it to be. It typically is a reflection of the interest you have in the project not the product. Most developers can hate what they work on and do incredible work if they love what they are doing minute to minute.
No I won't be a passionate evangelist about spreadsheets, but I could get happily lost in the codebase for Google Docs. People code for 48 hours when lost in the happy delirium of discovering some new problem, that's passion. However it might have nothing to do with their personal investment for the finished product. In fact they likely really enjoy the challenge of fixing something broken more than seeing it run once fixed.
Putting passion on a job spec is really quite unreasonable, but it seems easier than saying single minded to a fault and willing to sacrifice personal health to do a bit more hacking.
- I feel like teens are the stick used to beat Facebook because they've never really been the ones who fanatically used it.
This study seems to constantly come out saying teens are using x messenger app. Facebook isn't a messaging system like Snapchat or WhatsApp. It's a repo for social contacts and experiences, which is why it has a symbiotic relationship with messaging apps (people sent texts while Facebook was only on the web! Facebook is losing ground to texts!).
Of the four Twitter and Instagram are the competitors but both lack the walled garden that Facebook offers. Facebook for all it's privacy concerns makes these look like a safe-house.
Instagram treads the role of messaging app and social network best, and is owned by Facebook and is deeply integrated. That doesn't seem like a dead and buried social network.
- Cities that don't care for their homeless tend not to have a homeless problem.
SF actually has relatively great care for homeless people compared to the majority of the US. Therefore a lot more homeless end up there.
The homeless problem is striking because it's almost entirely artificial. If SF didn't care for it's homeless, and didn't rent control areas to allow affordable housing it would probably be like midtown manhattan. That's boring to a lot of residents and unfortunate for the homeless, but where's the tax money coming from to allow the status quo to continue?
No one seems to be debating the value of social programs, but what's the advantage of the tenderloin? It's been a decaying mess for generations when it could be a profitable, vibrant part of the city which would be beneficial for wealthy and poor alike.
- Javascript having a higher approval rating than coffeescript is actually pretty interesting.
- Nothing like a phobia or anything, but it did give me a feeling that there were bugs all over my screen.
It's a cool animation though and a nice reference to the first book, personally I'd probably add a small toggle for the bugs.
But I think Open AI is not a slam dunk for Ads. Gemini and AI mode will compete for the same budget, and Google's Ad machine is polished.
I think eventually you will buy Ads for Open AI in Google's marketing platforms, just like most people buy bing ads in Google.