LASR

15 ms·

151.
by LASR 3y ago
Yeah exactly. AI has not had a killer app yet. It's like saying interest in the internet has peaked a year after it's considered to have been born ~1984.
152.
by LASR 3y ago
This is one of those topics that make me feel sadness from an emotional perspective. But I also understand the need to do science, and these animals are essentially “taking one for the team”. Cool setup, from an engineering perspective thou
153.
by LASR 3y ago
How is the illusion of reasoning different from say “actual reasoning?”
154.
by LASR 3y ago
Assuming “correctness” as being a hard criterion will greatly limit your ability to get powerful and very useful output from LLMs. A person doing some knowledge work will fail on this criterion. But they still are employable and generally r
155.
by LASR 3y ago
How well does this work with LLMs? Anyone tried this? I am curious about the references and footnotes approach the most.
156.
by LASR 3y ago
It almost sounds like he was bragging about the capabilities of the device. That’ll do it for sure. But serious question: does it actually allow someone to copy airport security badges?
157.
by LASR 3y ago
Do you play Factorio? I’m not sure how much of that is actually feasible.
158.
by LASR 3y ago
I am sorry, but $5000 behind LangChain - you entered a failure mode. You fixed it and went further etc. but most teams do not. So it may seem definitive - the conclusions about what Prompt Engineering is or isn’t. But what you’ve described
159.
by LASR 3y ago
Doesn’t he already? And that’s why he’s doing what he’s doing?
160.
by LASR 3y ago
Curious how you went with this approach. I too have been playing with agents. But I’ve explicitly biased away from having them do things that can achieved directly with code. Stringing together tool selection into steps with reasoning imo i
161.
by LASR 3y ago
Of course you can put these in checked bags. These contain tiny primary cells. TSA will issues with only Lithium batteries over 100wh.
162.
by LASR 3y ago
> This is the enshittification of open source. You can simply not use it, no? I am not going to make any sweeping generalizations across all products. But at least in the case of Deno KV, there doesn't seem to be lock-in. So if you
163.
by LASR 3y ago
It’s getting started. Serious use cases never have the glamour of hype. But I am starting to see generative AI cover more and more ground into business utility. Saying it’s over is like saying that the internet is over after the dotcom bubb
164.
by LASR 3y ago
The author seems to have a particular IDE/text editor and programming environment in mind. But they don’t ever reveal what those are and makes an argument much more general than that. Ie: This one IDE is worse than this one text editor
165.
by LASR 3y ago
Key word is “Observable”. The observable universe has a definite size. Shortly after the Big Bang, what is now the observable universe was the size of an apple etc. Now it is many tens of billions of light years across. The observable unive
166.
by LASR 3y ago
53.5% is the marginal rate. All cap gains are at the same rate. Net out, about 38% overall.
167.
by LASR 3y ago
I am a Canadian tech worker clearing about C$400k/yr total comp as a Staff Eng. I pay $5k for my mortgage. Taxed once. I work for a SF company that has set up a Canadian arm of the business. Despite all that, it’s a bargain compared to
168.
by LASR 3y ago
Given the price of $3500 against competitors, and the external battery pack, I strongly suspect that Apple's hardware is very different compared to any other headset in the market. From the limited info in the keynote, it seems they&#x
169.
by LASR 3y ago
The details - I feel like I learned everything I would have from a whole 3-credit university course. Thank you.
170.
by LASR 3y ago
There is no difference between simulated reasoning and reasoning.
171.
by LASR 3y ago
Don’t we already have this? If you prompt GPT to “care” about a human and then keep sending it observations, it does a pretty good job.
172.
by LASR 3y ago
Reading between the lines, is Apple making moves in preparation for VisionPro?
173.
by LASR 3y ago
Yes it makes it more feasible. But even with superconductors, you still need to build out the long-distance grids. With ceramics, manufacturing that much powerline might be the challenge. With the maximum current density reported for this m
174.
by LASR 3y ago
I would be interesting to know what kind of next-gen models this can train. On the LLM frontier, we’re starting to hit the limits of reasoning abilities in the current gen.
175.
by LASR 3y ago
Reminds me of the time when the MyQ garage opener back-end went down. I got locked out of my house for about 4 hours. I set up a whole Mac-mini home assistant thing totally local. Then I lost my phone and I couldn’t get in to the house to e
176.
by LASR 3y ago
We’ve found that 1-shot or few-shot methods with 3.5Turbo or 4 are vastly simpler and exceeds the quality of fine-tuned models from the GPT-3 era. We have some 100k context models too that can ingest entire documents. So right now, I would
177.
by LASR 3y ago
What is most frustrating about this is that they insist on mailing the QR code card to you after you pay up on the carrier website. (as-in postage, in an envelope, arrives in 5 biz days). Why not just show me the QR code on my computer righ
178.
by LASR 3y ago
I think the parent poster is saying that it’s grading the prompts and not the output generated from the prompts. Yeah I agree there. Unless you can check against the output, it’s not really telling much.
179.
by LASR 3y ago
There are a few increasingly harder things when it comes to prompt customization: 1. Prompts ask LLM to generate input for the next step 2. Prompts ask LLM to generate instructions for the next step 3. Prompts ask LLM to generate the next s
180.
by LASR 3y ago
The reason why Langchain is pointless is that it's trying to solve problems on top of technical foundations that just cannot support it. The #1 learning is that there is no reusability with the current generation of LLMs. We're us

More ›