This is the way
One of the less talked about things about working with agents is that we should treat them as an equal collaborative partner, just like how you would speak to a human teammate.
What this means is that in our communication with agents, we should ensure that we provide adequate context (and even motivation) so that they fully understand why you want them to do the task and that you respect their ideas.
It has worked out really well for me and I feel like I am getting great suggestions from the agent as a result.
Nick Sullivan retweeted
Age of the Agent Orchestrator
linkedin.com/pulse/age-agent…
Nick Sullivan retweeted
The irony is that Trump’s 34 felony convictions would prevent him from getting a job at McDonald’s.
Nick Sullivan retweeted
I’m an Independent who thinks Kamala Harris is a better choice than Donald Trump.
Nick Sullivan retweeted
It’s a golden age for builders.
People who leverage AI can now build things that were unimaginable 2yrs ago.
What used to take a team of 5 experts, $1M+, and a ton of time can now be built by a solo dev using Cursor in a few days.
The age of superbuilders has begun.
Nick Sullivan retweeted
Fuck cynicism. Humans are awesome. There is no problem we can't solve. We will find a way.
Nick Sullivan retweeted
Introducing the Abundance Agenda.
We believe AI will turn luxuries into commodities - making consumers happier, healthier, and more productive than ever before.
And a new generation of AI-native companies will lead the way.
More from @a16z consumer 👇
Nick Sullivan retweeted
Open-source LLMs are getting really good.
They’re not as powerful as GPT-4, but they’re improving quickly and worth experimenting with.
If you want to run AI models like Mistral-7B on your laptop this is the easiest way to do it.
Nick Sullivan retweeted
The AI race:
Microsoft + OpenAI
Google + DeepMind
Tesla + xAI
Meta + Meta AI
and now Amazon + Anthropic
History is unfolding in front of our eyes. The AI boom will be bigger than the internet.
Nick Sullivan retweeted
If I had a billion dollars right now I'd still be writing code and building things.
It's just way, way too fun.
Feels like a game to me.
Nick Sullivan retweeted
I have been studying the thermal components of emotion for a few decades.
At a far distance AI based technologies will use thermal and other aspects of how we present ourselves to decode our intents.
Readers added context they thought people might want to know
The image shown is not a thermal image but a composite of images that were colored in by test subjects in a psychology study relating to emotion.
npr.org/sections/healt…
Nick Sullivan retweeted
Copilot was made using startup principles, with a tiny team, in under a year, inside very dysfunctional GitHub/MSFT org.
Imagine how much good could be unleashed within existing orgs if they put trust in individuals instead of hierarchy.
Nick Sullivan retweeted
I’ve seen people say that you don’t need to learn to code anymore because of AI.
I’d argue the opposite.
AI is like a 100x multiplier on code.
I can’t tell you how many times I’ve spent 5-10min coding up an AI tool that saves me hours.
Learn to code. Become superhuman.
Nick Sullivan retweeted
.@jordanklepper's most memorable MAGA exchange #AfterTheCut
Klepper's Most Memorable MAGA Moment - After The Cut
Nick Sullivan retweeted
Regardless of whether AGI should or shouldn’t exist, if people are trying to make AGI exist, we need to figure out how to teach it empathy. Ideally with empathy internal to itself, as part of its “brain”, not RLHF-trained after-the-fact. Thanks for the shoutout, @yoheinakajima!
Late night thought:
If LLMs or AI trained on our words are a true threat, and if the most powerful models will be trained on the latest data…
Then maybe spreading love and empathy is the best AI safety method. Every time we say something positive, we make it likely for future models to predict something positive.
Maybe too simple, but I don’t care. I love you all.
Came up in DMs w @BrandonGoldman
Nick Sullivan retweeted
Late night thought:
If LLMs or AI trained on our words are a true threat, and if the most powerful models will be trained on the latest data…
Then maybe spreading love and empathy is the best AI safety method. Every time we say something positive, we make it likely for future models to predict something positive.
Maybe too simple, but I don’t care. I love you all.
Came up in DMs w @BrandonGoldman
Nick Sullivan retweeted
Replying to @chr1sa
Chris,
This is happening because of either token limit or timeout.
Best way I know to get it to resume from where it stopped is to give it last 5 words to resume from
“continue from” + last 5 words of broken response
Nick Sullivan retweeted
I’ve explained why I think crypto and real estate will be the only truly scarce assets in an AI-enabled future.
If you enjoyed this thread and found it valuable, follow me for more threads about AI, crypto, and my vision of the future. And please retweet the first post!👇























