haltist

17 ms·

301.
by haltist 3y ago
Do you build competitive AI models?
302.
by haltist 3y ago
The only metric that matters for a business is whatever helps their bottom line. No one really cares about competitive results if they can just fine tune some open source model on their own data set and get good business outcomes. So if the
303.
by haltist 3y ago
What's hard about AI that requires special expertise? In many ways it is much simpler than regular software engineering because the conceptual landscape in AI is much simpler. Every AI framework offers the same conceptual primitives an
304.
by haltist 3y ago
I've been walking every day between 6-10 miles depending on weather for the last 3 months and I have gotten very good at walking, running, and sprinting. I am starting to do some basic agility training like parkour as well. I'll p
305.
by haltist 3y ago
If it was 20 years ago then you probably don't have a brain tumor because those are caught very late and are often terminal. Mine is from some high impact sport injuries so there is no permanent solution but if someone doesn't hav
306.
by haltist 3y ago
It is possible by reducing neck and jaw tension but there is no permanent solution. Reducing caffeine intake also helps because caffeine is a stimulant and makes muscle tension worse.
307.
by haltist 3y ago
Something else to consider is that if it is possible to truly engineer human level intelligence for operating autonomous machines then ethical concerns would require treating such systems with full human rights. Some are already claiming th
308.
by haltist 3y ago
It should be possible to embed cubical type theory in Idris.[1][2] 1: https://arxiv.org/abs/2210.08232 2: https://chat.openai.com/share/aadb7a0a-08a4-4951-b877-cb2f61...
309.
by haltist 3y ago
Same ideas have been applied in data centers to minimize energy use for dissipating excess heat from server racks. In theory, AI should be able to optimize any kind of "plant" and its processes to minimize some metric like energy.
310.
Introduction to Meta-Learning (2020)(pub.towardsai.net)
2 points by haltist 3y ago|0 comments
311.
by haltist 3y ago
How much was the CEO getting paid for making stupid decisions?
312.
Companies are making serious money with AI(sloanreview.mit.edu)
2 points by haltist 3y ago|2 comments
313.
by haltist 3y ago
Would be interesting to see if their work discovers any prime concepts of kinetics (static and dynamic).
314.
Self-Taught Optimizer (Stop): Recursively Self-Improving Code Generation(arxiv.org)
1 points by haltist 3y ago|0 comments
315.
by haltist 3y ago
Another fun project would be conspiracy theory books and articles.
316.
by haltist 3y ago
Someone should train an LLM on handbooks of rationality to see what happens.
317.
Understanding LLMs Through the Problem They Are Trained to Solve(arxiv.org)
4 points by haltist 3y ago|0 comments
318.
by haltist 3y ago
Very interesting concept.
319.
Large Language Models Fail on Trivial Alterations to Theory-of-Mind Tasks(arxiv.org)
3 points by haltist 3y ago|0 comments
320.
by haltist 3y ago
You can try the experiment yourself. Take any open source LLM and then feed the output back into it in a loop and you'll see what I mean. Most LLMs diverge and lose coherence in less than 100 iterations.
321.
by haltist 3y ago
In system design there is something called resonant amplification and what you are describing is very similar. The biases of the model are amplified with each iteration and the end result is that the system converges onto the patterns recog
322.
by haltist 3y ago
I am assuming humans don't go extinct but if the global economic system collapses then yes, the 3rd option is essentially extinction or at the very least an extremely drastic reduction of the human population.
323.
by haltist 3y ago
It's hard to grasp that the earth is a system within a finite volume of space. I agree it is obvious but very few people actually understand what it means to live within a bounded volume of space with finite resources.
324.
by haltist 3y ago
Children are now born with plastic in their veins. We are already at the limits of growth and potentially beyond it so either some technological miracle enables further growth or we start facing the negative consequences of cancerous growth
325.
by haltist 3y ago
The biomass of earth is finite so the neoliberal economic directive of constant exponential growth is both a physical and logical impossibility. Either the economic model changes and we start consuming within bounds that helps the earth sys
326.
by haltist 3y ago
For practical purposes it doesn't really matter because most real world computation is an approximation anyway. AI researchers don't care much about impossibility results because their goal is human level artificial intelligence a
327.
by haltist 3y ago
Pratt gives a very good answer with a historical overview of the main characters and their theorems on incompleteness: https://lists.ugent.be/wws/arc/fom/2023-10/msg00007.html
328.
by haltist 3y ago
Why don't large tech companies invest in modular nuclear reactors? It's an obvious solution to their energy consumption problems. Solar would work as well but it doesn't have the required energy density for powering their com
329.
Ernest Rutherford(history.aip.org)
1 points by haltist 3y ago|0 comments
330.
by haltist 3y ago
The root cause of the problem is energy expenditure > How does that happen? Look at the collapse in food per capita. Right now, every calorie we eat actually contain up to 10 kcal worth of fossil fuel energy. This is an energy return on

More ›