🥋

fixtures/bensbites.beehiiv.com/19-insights-microsofts-future-work-report-2023.md

@@ -0,0 +1,53 @@
+# 19 insights from Microsoft's Future of Work Report 2023
+
+Microsoft released their annual [Future of Work report](https://www.microsoft.com/en-us/research/uploads/prod/2023/12/NFWReport2023_v4.pdf?utm_source=bensbites\&utm_medium=referral\&utm_campaign=19-insights-from-microsoft-s-future-of-work-report-2023) and this time around it’s not about remote work, it’s about AI. Like no one would guess that.
+
+The report has stats from many studies done in 2023, backed by theoretical research from past years. I compiled what you want to know in this busy man’s guide.
+
+*(Disclaimer: some stats are approximated, and takeaways paraphrased. Slide numbers are in parentheses after each bullet point, refer to the the actual report if you want to cite stuff).*
+
+![](https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/6cb186c7-06a5-4da4-b88d-8e1a186ee313/image.png?t=1706534618)
+
+1. Knowledge workers with ChatGPT are **37% faster, 40% higher quality but ~20% less accurate.** Simple UX solutions to solve this are possible. `(6)`
+
+2. From a survey of enterprise users of Microsoft Copilot 365 `(7)`:
+
+   - **73% agree that Copilot makes them faster**
+
+   - 85% said it would help them get to a good first draft faster.
+
+   - 72% agreed about spending less mental effort on mundane or repetitive tasks.
+
+3. Most early studies have found that new or **low-skilled workers benefit the most** from LLMs. Less skilled workers improved by 43% vs more skilled who improved by about 17%. `(8)`
+
+4. Assistant needs to be paired with provacators i.e. LLM-based tools that challenge assumptions, encourage evaluation, and offer counterarguments. `(9)`
+
+5. AI can help with breaking down simple commands into micro-moments and microtasks, improving overall quality and efficiency. `(10)`
+
+6. Analyzing and integrating AI-generated information may become more important than searching and creating information. Skills not directly related to content production (leading, social interactions, trust issues, or emotional awareness) may be more valuable. `(11)`
+
+7. **Prompting is hard, but people are getting good at it.** Fine-tuning/using LLMs to generate prompts is making it easier as well. Prompt templates are helpful for end users. `(12-14)`
+
+8. Highlighting errors/uncertainty percentages can help balance reliance on LLMs. Prompting can be complemented with co-audit tools to check LLM outputs. `(17-18)`
+
+9. Generative AI requires self-awareness and well-calibrated confidence. At the same time, it can help in getting there too. `(19)`
+
+10. Creative activities are a process and LLMs can help across different parts. `(21)` 69% of Bing Chat conversations are in domains oriented toward professional tasks. `(22)`
+
+11. **A larger chunk of LLM-based searches is complex** (36% of them) than traditional searches (13% are complex). `(22)`
+
+12. In a study of 69 students, the use of Codex improved their performance in learning Python, but it did not impact their manual code-modification abilities. `(24)`
+
+13. LLMs can rapidly analyze data from humans and generate synthetic data. That’ll **change how social science research is done.**`(27)`
+
+14. LLMs in meeting can solve different problems like equal participation (instant feedback) and better interactions (retrospective feedback) `(28-29)`.
+
+15. AI can help in delegating management responsibilities, freeing execs to focus on team vision. `(30)`
+
+16. **Modern office knowledge is in chats**, not documents but applying AI over employee chats is tricky. `(31-32)`
+
+17. Approx. 80% of the US workforce could have at least 10% of their work tasks affected by GPTs. Around 19% of workers may have 50% of their tasks impacted. `(38)`
+
+18. “Innovation vs. automation” is often a better framework to use than “substitution vs. augmentation”. **Augmentation can still mean job loss.** It is important to try to track whether and where human labour is being used in innovative new ways. `(39)`
+
+19. Instead of “How will AI affect work?”, the question should be **“How do we want AI to affect work?**” `(40)`

fixtures/bensbites.beehiiv.com/ai-staff-google-now-betting-startup.md

@@ -0,0 +1,19 @@
+# AI staff from Google are now betting on their own startup.
+
+Heard about Google Deepmind, right? The team building Google’s AI models. [Three DeepMind researchers have decided to do their own thing.](https://www.theinformation.com/articles/more-google-deepmind-staff-depart-to-launch-an-ai-startup?utm_source=bensbites\&utm_medium=referral\&utm_campaign=ai-staff-from-google-are-now-betting-on-their-own-startup) They are joining the ever-increasing number of AI researchers who are moving out of Google. Meanwhile, Google's dangling special stock deals to keep their AI whizzes from walking out.
+
+## What’s going on here?
+
+AI staff from Google DeepMind are now betting on their own startup, fueled by millions in funding and big ideas.
+
+![](https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/5af98962-9ed0-45ac-b9e5-0134342aee04/image.png?t=1706094191)
+
+## What does this mean?
+
+The key players here are David Ding, Charlie Nash, Yaroslav Ganin, and maybe Conor Durkan too. These guys were behind DeepMind's Lyria and Imagen 2, models for creating tunes and pics. The startup is called Uncharted Labs and has already bagged $8.5 million (eyeing total $10 million). Big names like Andreessen Horowitz might chip in too.
+
+But why leave Google? Turns out, they got fed up with the slow pace and red tape. Google tried fixing things by merging their AI units, but that just ticked off the DeepMind crew even more. They felt like they lost their freedom to innovate. And it's not just them – other DeepMind and Google Brain talents are also jumping ship to start their own AI adventures.
+
+## Why should I care?
+
+Mistral is one of the leading players in the foundational models space just after few months after its launch. The key ingredient, if you ask me? Researchers from Meta and Google Deepming were the ones building. Character AI, Eleven Labs and a dozen others are similar. Plus, the venture capital folks are pouring cash into AI so I guess, more of these startups are coming our way in 2024.

fixtures/bensbites.beehiiv.com/apples-doubling-ai-siri-messages.md

@@ -0,0 +1,21 @@
+# Apple's doubling down on AI in Siri and messages.
+
+[Let's break down what's happening with Apple and AI.](https://9to5mac.com/2024/01/26/apple-siri-chatgpt-ios-18-development/?utm_source=bensbites\&utm_medium=referral\&utm_campaign=apple-s-doubling-down-on-ai-in-siri-and-messages) With its latest beta release iOS 17.4 update, we can see a glimpse of what's on the horizon for iOS 18. Apple's taking steps to revamp Siri and Messages with some testing against ChatGPT too.
+
+## What's going on here?
+
+Apple is expected to introduce advanced AI features in iOS 18, as suggested by iOS 17.4 beta code.
+
+![](https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/b4725d21-2e8e-4838-bc64-128708e39342/image.png?t=1706527756)
+
+## What does this mean?
+
+The code in this release shows a new feature called SiriSummarization to improve how Siri understands and summarizes information, potentially making it more useful in daily tasks and communication. ChatGPT is lending a hand for now, but Apple is also testing its own AI models, such as "Ajax," to power these improvements.
+
+Bloomberg also reported potential AI integration in Messages for auto-responses and sentence completion, driven by Craig Federighi's software team.
+
+## Why should I care?
+
+In the iOS 17.4 beta, there's clear evidence that Apple is experimenting with enhancing Siri's capabilities, but it’s unlikely that Apple will use external models. This is more like testing their internal models vs ChatGPT (and Flan-T5 too).
+
+Given Siri's (and honestly, most voice assistants’) lacklustre performance up until now, this move by Apple could be a game-changer. We all want assistants that can actually understand context and provide relevant, useful interactions. Let's see if Apple can take the lead in making assistants truly assistive.

fixtures/bensbites.beehiiv.com/bret-taylors-stealth-ai-startup-almost-unicorn.md

@@ -0,0 +1,21 @@
+# Bret Taylor’s stealth AI startup is almost an unicorn.
+
+[Bret Taylor’s AI startup is about to become an unicorn.](https://www.bloomberg.com/news/articles/2024-01-26/bret-taylor-s-ai-startup-to-get-funding-at-near-1-billion-value?utm_source=bensbites\&utm_medium=referral\&utm_campaign=bret-taylor-s-stealth-ai-startup-is-almost-an-unicorn) The ex-CEO at Salesforce and an OpenAI board member, is making moves with his AI firm, Sierra.
+
+## What’s going on here?
+
+Sierra is in talks for new funding valuing it at $1B.
+
+![](https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/e2fb5904-33b3-4d68-9a2e-39ad2ea2f637/image.png?t=1706272109)
+
+<small>via Bloomberg</small>
+
+## What does this mean?
+
+Taylor founded this AI startup with Clay Bavor, a former Google hotshot. Now, it is on the brink of hitting unicorn status with a cool nearly $1 billion valuation, thanks to Sequoia Capital leading a hefty $85 million investment round.
+
+Benchmark led an early investment in the company last year and Sierra already has some significant traction. With big names like Taylor and Bavor at the helm and Sequoia Capital opening its wallet, Sierra's one to watch. Taylor has already said that he’s with OpenAI for a short while—until the board situation sorts out. He’s got plans.
+
+## Why should I care?
+
+Taylor’s got a knack for leading and innovating—riding Salesforce's ups and Salesforce's downs, and now trying to stabilise OpenAI’s board. Such people can’t sit on the sidelines when AI is changing industries. Together with Bavor, Taylor is bringing that magic to Sierra.

fixtures/bensbites.beehiiv.com/clearbits-team-uses-ai.md

@@ -0,0 +1,90 @@
+# How Clearbit’s team uses AI
+
+### and how you can implement similar strategies at work
+
+![Author](https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/user/profile_picture/fc858b4d-39e3-4be1-abf4-2b55504e21a2/thumb_uJ4UYake_400x400.jpg)
+
+[Ben Tossell](https://bensbites.beehiiv.com/authors/fc858b4d-39e3-4be1-abf4-2b55504e21a2)\
+January 25, 2024
+
+👋*Hey, this is Ben with a* ***🔒**\*\*subscriber-only issue* ***🔒**\*\*of Ben’s Bites Pro.* *A weekly newsletter covering how businesses are using AI.*
+
+*If you’re not a subscriber, here’s what you missed recently:*
+
+- *[How Deel uses AI in its business](https://bensbites.beehiiv.com/p/deel-uses-ai-business)*
+
+- *[How Sandwich Video uses AI internally](https://bensbites.beehiiv.com/p/sandwich-video-uses-ai-business)*
+
+- *[AI’s impact on jobs: A look at front-end engineering](https://bensbites.beehiiv.com/p/ais-impact-jobs-look-frontend-engineering)*
+
+*Subscribe to access these and all future & past posts.*
+
+[<button>Subscribe to PRO</button>](https://bensbites.beehiiv.com/upgrade)
+
+I’m exploring how to implement AI in business and today I’m excited to show you how Clearbit does it.
+
+[Clearbit](https://clearbit.com/?utm_source=bensbites\&utm_medium=referral\&utm_campaign=how-clearbit-s-team-uses-ai) provides company data to sales and marketing teams to find and engage their ideal customers. They were acquired by HubSpot in November 2023 for $150M.
+
+Thank you, [João Moura](https://twitter.com/joaomdmoura?utm_source=bensbites\&utm_medium=referral\&utm_campaign=how-clearbit-s-team-uses-ai), Director of AI Engineering at Clearbit, for spending time with me and opening up behind the scenes. He also shares his AI agent framework, [CrewAI](https://www.crewai.io/?utm_source=bensbites\&utm_medium=referral\&utm_campaign=how-clearbit-s-team-uses-ai).
+
+What stood out to me from this conversation:
+
+- exploring use cases with an AI champion in charge
+
+- rolling it out to the team by forming small, cross-functional teams and speaking with external experts
+
+- the challenges you’ll face when doubling down on AI implementation in your product and team
+
+- which tools they use and why
+
+- leveraging AI agents to automate your life and get time back
+
+- how much AI has changed the way different teams work
+
+- steps for AI adoption in your team
+
+## The exploration phase
+
+When João first brought up AI into Clearbit, it seemed obvious. They have all this data about so many entities, what can they do with it?
+
+So Clearbit began experimenting with AI, and João developed a predictive model for the likelihood of winning sales opportunities, which could help customers forecast their deal pipelines using their own and Clearbit’s data.
+
+They also started building a GPT plugin and using LLMs for improved data parsing, as much of their work involves extracting internet data gems.
+
+It basically integrated with some of our core APIs, it would allow Clearbit customers to pull data on the website visitor, search for similar companies and do some early prospecting. The good thing was that by getting that data into the Chat you could easily translate from checking visitors to finding similar ones and to asking it to help you write emails.
+
+This quickly evolved into complex use cases with RAG systems and large-scale embeddings.
+
+Retrieval-augmented generation (RAG) is like a smart assistant that can look up information and talk about it naturally, while Regular Expressions (Regex) is a tool for finding specific words or patterns in a text.
+
+RAG models are really good at figuring out the context to answer questions or create text, using lots of information beyond what Regex can do with its basic pattern finding, making them more suitable for complicated language tasks.
+
+Before there were only so many cases you could cover with Regex, with RAG we can tap into data points we wouldn't be able to easily extract otherwise.
+
+Embeddings turn words or items into numbers so that computers can understand how similar or different they are to each other.
+
+Embeddings were crucial to achieving effective RAG because they provide a nuanced understanding of language that is essential for both the retrieval and generation components of RAG models, like tagging and classification.
+
+Embeddings were intrinsical to get some good RAG, but we also have leveraged them to do actual tagging and classification in a scale and level of precision we wouldn't be able to before by relying only on ingesting data from sources
+
+AI unlocked numerous use cases, diverse data sources, and processes, enhancing data coverage and quality, making their company data superior to any in comparison.
+
+As of today we now have 100% coverage across three of the most important company data points. Every single company domain requested by a customer within the last 3 months (~4M) has a clear english description, accurate industry categorization, and detailed company tags.
+
+Matt Sornson - GM & VP of Product
+
+They were able to infer data and tag companies with higher precision and that they wouldn't have been able to in the past.
+
+Take a look at the % increases (in green) below for their metrics:
+
+![](https://lex-img-p.s3.us-west-2.amazonaws.com/img/75a621c7-f951-4aa4-b163-bf66eee81cad-RackMultipart20240125-189-v51zbs.jpeg)
+
+## Rolling out AI to the team
+
+## Subscribe to Ben's Bites Pro to read the rest.
+
+Become a paying subscriber of Ben's Bites Pro to get access to this post and other subscriber-only content.
+
+[Upgrade](https://bensbites.beehiiv.com/upgrade)
+
+Already a paying subscriber? [Sign In](https://bensbites.beehiiv.com/login)