__alexs

8 ms·

91.
by __alexs 6mo ago
I feel like most engineers I talk to still haven't realised what this is going to mean for the industry. The power loom for coding is here. Our skills still matter, but differently.
92.
by __alexs 6mo ago
I miss my Treo :(
93.
by __alexs 6mo ago
Another weird cultural quirk of the Dutch that will hopefully go the way of Zwarte Piet one day.
94.
by __alexs 6mo ago
Anthropic's marketing team are terrifyingly good. I wonder if Opus came up with this plan?
95.
by __alexs 6mo ago
It cannot do "anything" with the tools. Tools are very constrained in that the agent must insert into it's context the tool call, and it can only receive the response of the tool directly back into its context. Tools themselv
96.
by __alexs 6mo ago
The agent cannot compose MCPs. What it can do is call multiple MCPs, dumping tons of crap into the context and then separately run some analysis on that data. Composable MCPs would require some sort of external sandbox in which the agen
97.
by __alexs 6mo ago
This hasn't been the full story for years now. All SOTA models are strongly post-trained with reinforcement learning to improve performance on specific problems and interaction patterns. The vast majority of this training data is gene
98.
by __alexs 6mo ago
I think OpenAI and Anthropic probably have a lot of that lying around by now.
99.
by __alexs 6mo ago
The models are already massively over trained. Perhaps you could do something like initialise the 2 new token sets based on the shared data, then use existing chat logs to train it to understand the difference between input and output cont
100.
by __alexs 6mo ago
Why are tokens not coloured? Would there just be too many params if we double the token count so the model could always tell input tokens from output tokens?
101.
by __alexs 6mo ago
Solving arbitrary logical problems seems to be equivalent to solving the halting problem so you are probably wise not to make that bet.
102.
by __alexs 6mo ago
A lot of GPUs in this list are basically just previous GPU but faster or more RAM. I kind of thought it was going to focus on interesting new architecture innovations.
103.
by __alexs 6mo ago
If you know even the basics of ethics then such claims are clearly nonsense. There is no stable context independent ethical behaviour. This is a great example of the dangers of motivated reasoning.
104.
by __alexs 7mo ago
There have been other 4D video games with actually good performance though.
105.
by __alexs 7mo ago
No it isn't? Are you an AI?
106.
by __alexs 7mo ago
Many people seem to believe the Claude Code has some sort of secret sauce in the agent itself for some reason. I have no idea why because in my experience Claude Code and the same models inside of Cursor behave almost identically. I think a
107.
by __alexs 7mo ago
I thought they needed $7 trillion or they'd be unable to keep training new models?
108.
by __alexs 7mo ago
Looking forward to someone patching it so that it works with non Anthropic models.
109.
by __alexs 7mo ago
Using some ML to derive a sentiment regex seems like a good actually?
110.
by __alexs 7mo ago
Not a strong indicator that Enterprises have good taste though is it?
111.
by __alexs 7mo ago
The reset isn't the problem, the entirely nerfing the Red team is the problem. The US took steps to fail to learn from the exercise before it had even finished.
112.
by __alexs 7mo ago
No they were slow at doing features before, and they are still slow afterwards.
113.
by __alexs 7mo ago
It's not cynical it is materialism. Shoppers do not want to pay to shop. Retailers pay thousands to encourage you to shop with them. They are the economic buyers of this feature.
114.
by __alexs 7mo ago
GitHub have not really got much better at adding new features either though :(
115.
by __alexs 7mo ago
The buyer of this technology is not shoppers, it's retailers. The measurement of quality is "does it make us more money?" not "does it help me make better buying choices." Retailers do not want you to make better ch
116.
by __alexs 7mo ago
(Good) E-commerce has been ruthlessly optimised to get shoppers to products they'll actually buy and then remove all distractions from buying. A chat interface is just fundamentally incompatible with this. The agent makes it too easy t
117.
by __alexs 7mo ago
Advertising your model with some obviously home grown benchmark is a bold play. It doesn't matter how good your model is, I immediately trust it less.
118.
by __alexs 7mo ago
They are also weirdly bad at Brainfuck which is basically just a subset of C.
119.
by __alexs 7mo ago
I had hope we might finally be ushering in a bold new era of programming in Malbolge but apparently that was too optimistic.
120.
by __alexs 7mo ago
Yes and LLMs can iterate and help you refine that spec more quickly.

More ›