sharemywin

6 ms·

1.
by sharemywin 21d ago
it would be cool to use embeddings to categorize into like 10-20 categories and get a giant pie chart
2.
by sharemywin 21d ago
at least now we can put a price on the cost to harm children.
3.
by sharemywin 21d ago
I feel like this might be a bot account. you posted no submissions, no favorites and your comments have a pattern. they're short, negative and not much variation.
4.
by sharemywin 21d ago
because code is only a small part of software development. don't forget ping pong tables. lol. I'm both making a valid point and making fun of it too.
5.
by sharemywin 21d ago
I thought about something like this I also thought about an affiliate program for it where people got there own link and got paid 50% of the ad sale if someone bought.
6.
by sharemywin 21d ago
Gee, wonder if this will result in a slap on the wrist and lots of cash...
7.
by sharemywin 2mo ago
that's the trillion dollar questions though can you teach it the expertise or does learning the expertise take 5,10,20 years to learn and what happens if things change does it do better or worse. how much if it's how fast an exper
8.
by sharemywin 3mo ago
a large % of wealth is a made up number. if everyone runs to the door to sell an asset it becomes close to worthless pretty quickly.
9.
by sharemywin 3mo ago
because the economy thrives on surprises
10.
by sharemywin 4mo ago
That suggests a better architecture for business processes: AI interprets. Rules validate. Humans approve exceptions or high-risk actions. Deterministic systems execute and record. right on the money. although I would say AI could possibly
11.
by sharemywin 4mo ago
so the earth is flat
12.
by sharemywin 4mo ago
that's kinda cool.
13.
by sharemywin 4mo ago
a little late to the party right?
14.
by sharemywin 5mo ago
good thing the current administration was basically elected to fix it. how's that going?
15.
by sharemywin 5mo ago
you might look at this for feature ideas. https://www.reddit.com/r/selfhosted/comments/1lutdul/we_buil... tldr; We built an open-source, MIT-licensed PaaS that: Lets you scale beyond a single server. Use
16.
by sharemywin 5mo ago
maybe robot.txt should be upgraded with license specifics. not for commercial use, etc. so I believe all content is covered under fair use which to me means common crawl has a right to scrape everything and it's the user of common craw
17.
by sharemywin 6mo ago
VCs now that AI will solve everything don't need too worry about climate change!!?!
18.
by sharemywin 6mo ago
Not bad for about $12-$16B in total actual revenue. net income probably: $1.5B – $3B P/E:500-1000 Of course people will trip overthemselves to buy it up.
19.
by sharemywin 6mo ago
I could see this playing out the that way
20.
by sharemywin 6mo ago
at least the 3-card Monty guy is dealing with you one on one. These companies hire lobbyist to make the illegal, immoral sh*t legal. Easy to keep winning the game when you write the rules.
21.
by sharemywin 6mo ago
but this is the opposite right? they own the customer relationship. Amazon does the opposite. They control the customer relationship. Can the supplier raise prices possibly but so can they middle man. if they turn over the relationship to t
22.
by sharemywin 6mo ago
and you don't even need to hold them until friday. at the very least sell your bullish position on thursday nights.
23.
by sharemywin 6mo ago
I guess I don't understand. so how does the agent know which tools to call versus an mcp?
24.
by sharemywin 6mo ago
I wish there was a company that was easy to use but wouldn't sell out in this arena.
25.
by sharemywin 6mo ago
they should give voting stock out as punishment.
26.
by sharemywin 6mo ago
climate change was an important issue when they were trying to peddle EVs and solar.
27.
by sharemywin 6mo ago
thought of making it an mcp server
28.
by sharemywin 6mo ago
if you want a job then here's my 2 cents: To me the opportunity is with agents. Specially copilot and what ever amazon's agent it. figure out how to code using them. build something cool in the space your interested in finding a j
29.
by sharemywin 6mo ago
People assume training on past data means no novelty, but novelty comes from recombination. No one has written your exact function, with your exact inputs, constraints, and edge cases, yet an LLM can generate a working version from a prompt
30.
by sharemywin 6mo ago
Wonder how much WFH could help. Seems like during covid demand went way down.

More ›