Gillesray

5 ms·

1.
Should Sam Altman fear token compression?
6 points by Gillesray 7mo ago|0 comments
2.
Should Sam Altman fear token compression?(edgee.ai)
6 points by Gillesray 7mo ago|1 comments
3.
by Gillesray 7mo ago
AI bills are exploding despite costs dropping 10x every 18 months because every saving gets reinvested into more usage. BUT: - Today's prices are subsidized (OpenAI losing $14B in 2026, Anthropic burning $12B in one quarter, Cursor al
4.
You Can't Optimize What You Can't See. AI Cost Observability(edgee.ai)
1 points by Gillesray 7mo ago|1 comments
5.
Who's in Charge? Disempowerment Patterns in Real-World LLM Usage(arxiv.org)
2 points by Gillesray 7mo ago|1 comments
6.
by Gillesray 7mo ago
Although AI assistants are now deeply embedded in society, there has been limited empirical study of how their usage affects human empowerment. We present the first large-scale empirical analysis of disempowerment patterns in real-world AI
7.
by Gillesray 7mo ago
50% around coding, they nail the first usage
8.
by Gillesray 8mo ago
very interesting
9.
by Gillesray 8mo ago
indeed. It tells a lot of about token ecosystem
10.
Will I Be Paid in Tokens?(tomtunguz.com)
6 points by Gillesray 8mo ago|5 comments
11.
by Gillesray 8mo ago
Technology companies are adding a fourth component to engineering compensation : salary, bonus, options, & inference costs. Levels.fyi pegs the 75th percentile software engineer salary at $375k. Add $100k in inference & the fully lo
12.
AI economics looks like cloud's one(edgee.ai)
3 points by Gillesray 8mo ago|1 comments
13.
by Gillesray 8mo ago
Enterprise AI spending jumped 3.2× in 2025 to $37B (Gartner), while inference costs dropped 90%. The paradox isn't new, it's cloud economics on fast-forward. What took cloud 10 years to reveal is happening in AI in 2-3 years: a) L
14.
by Gillesray 2y ago
great sum up
15.
by Gillesray 2y ago
love it
16.
AI firms will soon exhaust most of the internet's data(economist.com)
4 points by Gillesray 2y ago|1 comments
17.
by Gillesray 2y ago
"One approach is to focus on data quality rather than quantity. ai labs do not simply train their models on the entire internet. They filter and sequence data to maximise how much their models learn. Naveen Rao of Databricks, an ai fir
18.
Largest GDPR Fines on Big Tech(techcrunch.com)
2 points by Gillesray 2y ago|0 comments
19.
Edgee demo day. See how to collect 100% of the data
1 points by Gillesray 2y ago|0 comments
20.
The Big Lie
3 points by Gillesray 2y ago|0 comments