alienbaby

8 ms·

151.
by alienbaby 10mo ago
``sometimes, it even happens with [antigravity]`` Isn't this a problem with the agent loop / structure, rather than the llm, in that case? The ide doesn't affect the models results, just what is done with those results?
152.
by alienbaby 10mo ago
Llm's do not verbatim disgorge chunks of the code they were trained on.
153.
by alienbaby 10mo ago
I'm curious in what kinda if situations you are seeing the model the do opposite of your intention consistently where the instructions were not complex. Do you have any examples?
154.
by alienbaby 10mo ago
I've wondered if it's because structured outputs rely on visual cues to impart meaning, and turning them into token streams looses that spatial structure.
155.
by alienbaby 10mo ago
Exactly what I thought and why I checked the comments , an lo and behold it's not just me..
156.
by alienbaby 10mo ago
I actually ended up enjoying reading the cards after the charts more than I did reading the charts, but the charts were really interesting too.
157.
by alienbaby 10mo ago
If it can consume public resources and has a vote, maybe?
158.
by alienbaby 10mo ago
I have agreed with this for at least a decade. Name your things in a way related to what they do. What does chef do? Garden? Pig? Burp? Nonsense.
159.
by alienbaby 10mo ago
This will be fine, until you can't tell the difference and they forgo the 'i asked'
160.
by alienbaby 10mo ago
This will be fine, until You can't tell the difference and people forgo the 'i asked' part.
161.
by alienbaby 10mo ago
There are good people in IBM. But they don't have the resources behind them anymore. Look at the market cap of ms, Amazon. Google, meta et al, compared to IBM.
162.
by alienbaby 10mo ago
They gave up on watson about 18 months before llm's popped up, and they have simply just not got enough cash on hand to compete. While the big boys grew fantastically bigger over the past 15 years as cloud happened ibm fumbled time aft
163.
by alienbaby 11mo ago
Ai is a lot more than ai generated media.
164.
by alienbaby 11mo ago
Did you see the reply from the author to similar statements?
165.
by alienbaby 11mo ago
You guys make me laugh.
166.
by alienbaby 11mo ago
This is speculation currently, the actual reason has not been determined
167.
by alienbaby 11mo ago
Interesting experiment, though I'm not certain quite how the models are usefully compared.
168.
by alienbaby 11mo ago
This was my exact first experience trying to build anything of significant size. Several projects later, I can work with the AI now to produce several thousand lines of code, I understand it, it is what I asked for, has the architecture I w
169.
by alienbaby 11mo ago
Can't they just ask for copies of the lyrics they are not allowed to use and s/lyrics//g the training set? I imagine the volume of text that will be removed would be relatively miniscule.
170.
by alienbaby 11mo ago
it sounded to me like an AI generated article covering it's ass, or at least written by someone who did naught but AI generated research as their own 'research'.
171.
by alienbaby 1y ago
Using ai does not absolve you from the responsibility of doing it correctly. If you use ai, then you better have the skills to have done the job yourself, and so have the ability to check the AI did things correctly. You can save time still
172.
by alienbaby 1y ago
Until agi can sit there and ponder its own existence of is own violition and has the means to act upon it's conclusions, I'm not too worried.
173.
by alienbaby 1y ago
Can I just start using python if I've already got a bunch of projects manage with venv / pyenv / virtualenv ( and tbh I've kinda got into a confused mess with all these venv things, and at this point just hope they al
174.
by alienbaby 1y ago
I'm not a fan of being asked for payment before I know what I'm getting
175.
by alienbaby 1y ago
Nice. I wrote something exactly like this to simulate a CPU with an instruction set and registers that were extended beyond general purpose instructions to include op codes to read board state and generate moves for a game of go. That then
176.
by alienbaby 1y ago
ok, but if you did this in an inteview I was giving I would say "clever, but you didn't get the job". Mainly because you missed the point of the interview. We have a guy who joined recently who constantly tries to be as clev
177.
by alienbaby 1y ago
reminds me alot of patterns from diffusion limited aggregation.
178.
by alienbaby 1y ago
I thought I had learned a new word reading this, but instead I just have something that seems like it should be a word given the context it was discovered in. Perhaps that in itself should be considered cremement. A word that looks like it
179.
by alienbaby 1y ago
Hasn't Sam Altman already said they are profitable on inference, minus training costs?
180.
by alienbaby 1y ago
I'm not sure how running inference locally will make any difference whatsoever? or do you also mean hosting the MCP tools it has access to?

More ›