The rise of generative and AI large language models (LLMs) like ChatGPT
- Google drove a burst of innovation in the LLM space by sharing their knowledge and research with the AI world (Transformers)
- OpenAI, creators of ChatGPT, stole the LLM show. They made steady, solid progress over the last three years.
- Instead of doing their research, Microsoft invested early and hard in OpenAI ($1bn in 2019 and 2021), which paid serious dividends.
- Meta / Facebook also drove significant innovation in the field with LLaMa
- China is also making steady progress. A few popular LLMs include Wu Dao, PanGu-Sigma
- Amazon has steamed in at the end with Titan
- Complete data set here: https://docs.google.com/spreadsheets/d/1SEgPaYCckZFM5f7aGHbBqFN0IPMF2ajThvFUz2ap3bw/edit#gid=2145151656
Shopify offloads logistics business to Flexport
- Shopify is selling its logistics unit to the supply chain technology company Flexport. As part of the agreement, Shopify will receive stock representing a roughly 13% equity interest in Flexport.
- Shopify wants to go back to doing what we do best: building incredible software for e-commerce and letting Flexport do what they do best the logistics.
- Flexport will be Shopify’s official logistics provider. Shopify will also retain its Shopify Fulfilment Network app, where merchants manage their logistics process.
- Shopify president Harley Finkelstein said in an interview that after going on a “side quest” to develop the company’s own fulfilment and logistics businesses, it became clear that it could offer those services more effectively by integrating with Flexport.
Best practices are actually average practices.
Don’t be fooled by Best Practices. By the time something is labelled and advertised as a Best Practice, it is just average. Following these practices only suggests you won’t be left behind, not that you will lead the pack
Treat your to-read pile like a river, not a bucket
Treat your “to read” pile like a river (a stream that flows past you, and from which you pluck a few choice items, here and there) instead of a bucket (which demands that you empty it
Paul Graham on using AI to write
- When you lose the ability to write, you also lose some of your ability to think
- Observation suggests that people are switching to using ChatGPT to write things for them with almost indecent haste. Most people hate to write as much as they hate math. Way more than admit it. Within a year, the median piece of writing could be by AI.
- I [PG] often tell startups it’s safe to bet on laziness, and this is one of the biggest bets on laziness in history. The switch is going to happen, and we won’t know the consequences till it’s too late.
- Learn to use AI. It’s a powerful technology, and you should know how to use it. But also learn how to write.
Terrible data makes doing anything good impossible
Terrible data makes doing anything good impossible: Most organisations seem to have created their data stored as compacted Hell-bricks jammed into a teetering edifice which threatens to crush anyone who so much as breathes near it. It is frequently such a Herculean task to even put together a dataset that represents a clear view of what the organisation is doing that you just can’t.
Substack edition
The rise of generative and AI large language models (LLMs) like ChatGPT
Google drove a burst of innovation in the LLM space by sharing their knowledge and research with the AI world (Transformers)
OpenAI, creators of ChatGPT, stole the LLM show. They made steady, solid progress over the last three years.
Instead of doing their research, Microsoft invested early and hard in OpenAI ($1bn in 2019 and 2021), which paid serious dividends.
Meta / Facebook also drove significant innovation in the field with LLaMa
China is also making steady progress. A few popular LLMs include Wu Dao, PanGu-Sigma
Amazon has steamed in at the end with Titan
Complete data set here: https://docs.google.com/spreadsheets/d/1SEgPaYCckZFM5f7aGHbBqFN0IPMF2ajThvFUz2ap3bw/edit#gid=2145151656
Shopify offloads logistics business to Flexport
Shopify is selling its logistics unit to the supply chain technology company Flexport. As part of the agreement, Shopify will receive stock representing a roughly 13% equity interest in Flexport.
Shopify wants to go back to doing what we do best: building incredible software for e-commerce and letting Flexport do what they do best the logistics.
Flexport will be Shopify’s official logistics provider. Shopify will also retain its Shopify Fulfilment Network app, where merchants manage their logistics process.
Shopify president Harley Finkelstein said in an interview that after going on a “side quest” to develop the company’s own fulfilment and logistics businesses, it became clear that it could offer those services more effectively by integrating with Flexport.
Best practices are actually average practices.
Don’t be fooled by Best Practices. By the time something is labelled and advertised as a Best Practice, it is just average. Following these practices only suggests you won’t be left behind, not that you will lead the pack
Treat your to-read pile like a river, not a bucket
Treat your “to read” pile like a river (a stream that flows past you, and from which you pluck a few choice items, here and there) instead of a bucket (which demands that you empty it
Paul Graham on using AI to write
When you lose the ability to write, you also lose some of your ability to think
Observation suggests that people are switching to using ChatGPT to write things for them with almost indecent haste. Most people hate to write as much as they hate math. Way more than admit it. Within a year, the median piece of writing could be by AI.
I [PG] often tell startups it's safe to bet on laziness, and this is one of the biggest bets on laziness in history. The switch is going to happen, and we won't know the consequences till it's too late.
Learn to use AI. It's a powerful technology, and you should know how to use it. But also learn how to write.
Terrible data makes doing anything good impossible
Terrible data makes doing anything good impossible: Most organisations seem to have created their data stored as compacted Hell-bricks jammed into a teetering edifice which threatens to crush anyone who so much as breathes near it. It is frequently such a Herculean task to even put together a dataset that represents a clear view of what the organisation is doing that you just can’t.