Richard Skinner

Talk about learning and exploring Artificial Intelligence, machine learning and CTO stuff

[Richard Skinner @raskolnikoff101]

@fortelabs We built an app that does this for a client a while back.

[Richard Skinner @raskolnikoff101]

When applying our own agency we can achieve novel insights such as deep minds protein folding. My guess is that we will have to build agents that connect fast thinking models to reasoning models and apply other heuristics before we see novel autonomous breakthroughs. I’m not sure why we are expecting agency to be emergent.

[Richard Skinner @raskolnikoff101]

Brockmann, a font inspired by Josef Müller-Brockmann by atipo®. Download Medium weight for free!!! https://t.co/DADF8hOep9 via @atipostudio

[Richard Skinner @raskolnikoff101]

@SteveOnSpeed Good thing you weren’t working in a kebab shop

[Richard Skinner @raskolnikoff101]

@IrishTimes Did the copy editor for this title do their leaving cert English exam?

[Richard Skinner @raskolnikoff101]

@levelsio Tokyo and Kyoto have fantastic coffee scenes. Especially % Arabica, glitch, Fuglen, Koffee Mameya

[Ananay @ananayarora]

ChatGPT is down so people are going to Claude but Claude can’t browse the web so people are going to perplexity but perplexity is hitting chatGPT and GPT is down so it’s hitting Claude so with all this traffic chaos all 3 are now down. Also nobody uses Gemini but just to show they’re getting a lot of traffic they shut down too. That’s how we got the AI blackout

[Richard Skinner @raskolnikoff101]

@_ScottCondron @eugeneyan Thought I was missing something for a second, then I realized you weren’t being sarcastic.

[Richard Skinner @raskolnikoff101]

@jennyai I think that’s like saying 15 years ago that regular people aren’t using the cloud. This is a new tech stack not a new product. Regular people will use a new generation of products and not care what the tech is…

[RDI Hub @RDIhubIreland]

Introducing more speakers at our upcoming '#GenerativeAI for #IrishBusinesses' taking place Jan 16th in the #RDIhub: @raskolnikoff101, CEO of #Phased, with his topic 'AI Operations, Architecting the future of Enterprise Intelligence'
🎟️https://t.co/O3DPllpKSo

[Richard Skinner @raskolnikoff101]

@patrickc Would you not make the hardcover available in Ireland? 🇮🇪

[Richard Skinner @raskolnikoff101]

The challenges of integrating #GenerativeAI in organizations needs solutions in governance, data preparation and tech stack.

https://t.co/7B9PZPu143

[Richard Skinner @raskolnikoff101]

@jordantwestecom @nytimes do this with subscriptions

[Richard Skinner @raskolnikoff101]

@BBCWorld What is with the photos of Dublin in this piece? It’s like the photo journalist decided this was a daily mail style hatchet job on Ireland fueled by sour grapes and declining-empire indignation.

[Alex Ker 🔭 @thealexker]

The most interesting thing I've heard in a while:

Can AI ever be conscious?

Ilya Sutskever @ilyasut, co-founder at OpenAI, explained a simple experiment for AI consciousness.

We carefully curate the data such that we never mention anything about consciousness...we would only give the AI controlled data about surface-level experiences and mundane objects. At some point, we would tell AI about the concept of consciousness. If the AI agrees and says it's been experiencing the same thing or recognizes the experience, then it would be convincing.

[Richard Skinner @raskolnikoff101]

@sama This is your funniest tweet.

[Richard Skinner @raskolnikoff101]

@DrJimFan You still show up in the feed to blow away all the cringe. So it’s not all bad

[Richard Skinner @raskolnikoff101]

This is interesting.

[Anthropic @AnthropicAI]

Introducing 100K Context Windows! We’ve expanded Claude’s context window to 100,000 tokens of text, corresponding to around 75K words. Submit hundreds of pages of materials for Claude to digest and analyze. Conversations with Claude can go on for hours or days.

[Richard Skinner @raskolnikoff101]

@rachel_l_woods Surely this is just about vectorising different data sources such as internal slack channels, wiki etc. I think the challenge is finding an LLM that can be deployed locally that is anywhere near GPT-4.