ThomPete

5 ms·

1.
by ThomPete 6d ago
Once you realize that LLMs are great at style transfer a lot of things start to make sense and a lot of avenues open up for you. Also the "horrible" style is not because of AI it's because of what it's trained on. If any
2.
Show HN: Uplink – Software for Agents(uplink.spot)
2 points by ThomPete 11d ago|1 comments
3.
Uplink – Software for Agents(uplink.spot)
4 points by ThomPete 18d ago|0 comments
4.
Projects in Buzz(engineering.block.xyz)
2 points by ThomPete 1mo ago|0 comments
5.
Show HN: Block/buzz: a workspace built for teams of humans and agents(github.com)
21 points by ThomPete 3mo ago|7 comments
6.
by ThomPete 4mo ago
Network effect is the moat not domain expertise. Domain expertise build on network effect (for instance the knowledge of the network) yes, but just simple domain expertise is only a training round way.
7.
by ThomPete 4mo ago
That will fall again and everyone will be back to google
8.
by ThomPete 5mo ago
This is just flat out wrong. Making it removable means making it less effective, meaning using more materials etc. What is much more concerning is that you seem to be totally fine with the government deciding how something should be designe
9.
by ThomPete 5mo ago
So now the EU want to design phones too. What could possible go wrong.
10.
by ThomPete 7mo ago
same
11.
by ThomPete 8mo ago
Since it's inception in 1993, not a single global top 100 company have been created by any of the EU countries so the struggle is real. However I would strongly caution against thinking that reducing paperwork to set up your company is
12.
by ThomPete 1y ago
What revenue is that welfare state based on?
13.
by ThomPete 1y ago
Looks like it can
14.
by ThomPete 1y ago
Would it be possible to write a script for CAD that could do measurements
15.
by ThomPete 1y ago
It's really very simple. We used to have deterministic systems that required humans either through code, terminals or interfaces (ex GUI's) to change what they were capable of. If we wanted to change something about the system we
16.
by ThomPete 1y ago
My way out of this was to start thinking about what can't the LLMs of the world do and my realization was actually quite simple and quite satisfying. What LLMs can't replace is network effects. One LLM is good but 10 LLMs/age
17.
A brief history of the numeric keypad(doc.cc)
74 points by ThomPete 1y ago|19 comments
18.
by ThomPete 2y ago
Science do in fact rely on philosophy that's how we got the scientific method.
19.
by ThomPete 2y ago
Antelope AI | https://www.getantelope.com | Senior Python / Full stack developers | FULLTIME | ONSITE - HYBRID - REMOTE (US Only) | New York Antelope is the worlds first synthetic prediction market run by self-learning auto
20.
by ThomPete 2y ago
No the sun either is or isn't powered by nuclear fusion. There is no way of knowing whether that's the case until you can come up with a good explanation (hard to vary) everything before that is just guessing. I can assure you Deu
21.
by ThomPete 2y ago
Not from a non-bayesian perspective. It's what Deutsch would call a "bad explanation" i.e. it's easy to vary and thus doesn't tell us about how the sun is powered.
22.
by ThomPete 2y ago
Faktory | Full Time | Remote/On-site (New York, US) | https://www.faktory.com Faktory is building the worlds biggest, smartest and most reliable AI workforce. Your objective is to help make sure our ai agents become ever sm
23.
by ThomPete 2y ago
funny i get that too on twitter what you get on bluesky. if you seek out debate you will get it, has nothing to do with twitter. Bluesky will be very boring once upu realize that the reason you hate twitter is the reason you are going to ha
24.
by ThomPete 2y ago
This is crypto scare all over again. My answer is still the same you always have to do a cost benefit analysis. AI is many times more valuable than the energy it consumes and thus: AI doesn't have an energy problem, energy have an AI p
25.
by ThomPete 2y ago
Yes.
26.
by ThomPete 2y ago
Yes a few things. Distributed Prediction Market running purely on telegram: https://firstprinciple.co/popup/PopUp.mp4 Social Network based on interest using agents https://firstprinciple.co/popup/A
27.
Yale looking for 5 new fund managers to build fund with $50M each(investments.yale.edu)
1 points by ThomPete 2y ago|0 comments
28.
by ThomPete 2y ago
This is pure speculation and not really very well thought through IMO. We don't kill all ants either. Even humans work with the principle of collateral damage with each other so if the idea is that AI would suddenly be without mistakes
29.
by ThomPete 2y ago
"could" = "news making" not "news reporting" In other words this is pure speculation.
30.
by ThomPete 2y ago
visual control

More ›