JavaScript is not available.

May 16

Congratulations

@vipulved

@percyliang

and team

Quote Tweet

Together

@togethercompute

May 15

Great write ups on TechCrunch and Venture Beat covering our $20M seed funding led by @Lux_Capital. The stories include interviews with our founders and details not included in our blog post. Take a look! techcrunch.com/2023/05/15/tog venturebeat.com/ai/ai-startup-

May 14

One easy way to end up in this state is to optimize for CTR or engagement w/o normalizing for UI prominence

Looks like Twitter changed its recommendation function to indiscriminately boost videos ... Are others experiencing the same endless scroll of click-bait auto-playing popular and unpersonalized videos

The data + LLM space is going to get interesting soon Pressures: * Publishers want to get paid for the use of their data * EU / regulators * At some point, some big publisher will go after the open source datasets -- Common Crawl / Pile / Pushshyft / ...

May 13

All the things Google announced in I/O: - BARD improvements - PALM in BARD - PALM2 / ULM - MUM in Search - Gemini Branding advice -- simplify. Even if it's half a dozen codenames under the hood, brand them externally the exact same way. BARD1, BARD2, BARD3, BARD4, BARD5

Stack Overflow Will Charge AI Giants for Training Data

Apr 21

Another shoe drops ... wired.com/story/stack-ov Wonder how Reddit and Stack will draw the line between who pays and who doesn't? What about open-source LLM efforts like

GPT-J/GPT-Neo or

Alpaca or

Llama / OPT or

Dolly?

The programmer Q&A site joins Reddit in demanding compensation when its data is used to train algorithms and ChatGPT-style bots

Inside the secret list of websites that make AI like ChatGPT sound smart

Apr 19

Big news... 👾 #NeevaAI is officially available on

@poe_platform

! 🟪 Cited sources! 🟪 Up-to-date information! 🟪 Synthesized answers!

0:23

1.7K views

Concern trolling re: LLM training data from the Washington Post

washingtonpost.com

An analysis of a chatbot data set by The Washington Post reveals the proprietary, personal, and often offensive websites that go into an AI’s training data.

Reddit Wants to Get Paid for Helping to Teach Big A.I. Systems

The internet site has long been a forum for discussion on a huge variety of topics, and companies like Google and OpenAI have been using it in their A.I. projects.

Open web crawl always relied on a quid-quo-pro bt. aggregators and publishers ("you take my data; you send me clicks") With LLMs, the value exchange has tipped and pubs like Reddit go after cash-rich aggregators for their "fair share". May we live in interesting times ...

Apr 17

All the best. Truly impressive.

Quote Tweet

Elon Musk

@elonmusk

Apr 16

Launch attempt tomorrow

Excited to see

on the

AI 50 list!

Quote Tweet

Apr 12

Big news!

@Neeva is on this year’s @forbes AI 50! This list recognizes the most promising privately-held companies building businesses out of artificial intelligence.

Show this thread

Jack Clark

@jackclarkSF

Apr 10

Something that will further compound acceleration of AI research is that models are now better and cheaper than humans at data generation and classification tasks - this already true for some cases (eg Facebooks Segment Anything model, and GPT-4 for some labeling).

187

Yet another deep dive of how #NeevaAI works

Quote Tweet

Apr 5

Wonder how one might go about building a multi-doc summarizer? Want to know how it works under the covers? Read on for an overview of how our multi-doc summarizer (also known as #NeevaAI) is ahead of the game. Let's dive into it...

Show this thread

GIF

Fun with

AI! Ask the bard! Or speak genz to your kid!

Mar 31

Try #NeevaAI personalities to customize your

AI search experience ... What personalities would you like to see in #NeevaAI?

Quote Tweet

Mar 31

Watching our competitor chatbots hallucinate, fail at basic math, & fall in love — we felt like something was missing at @Neeva. That ends today with: #NeevaAI personalities

Feeling country?

On Bard's waitlist?

GenZ? Bet.

Captured by pirates? Follow along

Show this thread

0:50

1.8K views

It’s possible that large language models are more like the human brain than we thought, given that it is about prediction; that is why ChatGPT needs its own computer in the form of plug-ins.

Mar 29

Great article on ChatGPT plugins: stratechery.com/2023/chatgpt-l Everyone should be paying very close attention

stratechery.com

ChatGPT Gets a Computer

OpenAI: ChatGPT payment data leak caused by open-source bug

Mar 24

As I read about the OpenAI ChatGPT / GPT-4 history leak issue, I am not sure how this is not a bigger deal. bleepingcomputer.com/news/security/ At a company like

@Google

, this would be a "save all your docs, add the lawyers in and needs disclosure to the FTC" level privacy incident

bleepingcomputer.com

OpenAI says a Redis client open-source library bug was behind Monday's ChatGPT outage and data leak, where users saw other users' personal information and chat queries.

Mar 24

More thinking re: the OpenAI ChatGPT GPT-4 plugin launch ... The browsing plugin (WebGPT) uses Bing search. Bing just increased prices on that API 30x (from $7 for 1k requests to $200 for 1k requests) for AI applications OpenAI could not have pulled this off w/o MSFT ...

We’ve implemented initial support for plugins in ChatGPT. Plugins are tools designed specifically for language models with safety as a core principle, and help ChatGPT access up-to-date information,...

Mar 23

OpenAI GTM continues to be on a tear. Without looking at the details, just the sheer momentum of announcements and partnerships has me astounded

openai.com

ChatGPT plugins

Mar 23

Curious how to bake personalization into LLMs: 1/ Train a personal language model per user, e.g from flan-t5-xxl, or OpenChatKit's model and domain-specific pre-training on a user's own data (e.g Google Docs)? 2/ Shim of API calls to OpenAI with output massaging ... wdyt?

Pete Skomoroch

@peteskomoroch

Mar 22

The best time to invest in AI is 4-5 years ago 🚀

Adding the

gist widget to my iOS home screen was life-changing for me Follow the instructions in the thread below to do it ...

Quote Tweet

Mar 22

When you're on-the-go, you don't have time to waste scrolling through pages of search results. Stop searching & start discovering on @Neeva Gist, powered by #NeevaAI. And now available as a:

Widget on your home screen

Siri integration Let's get into the nitty gritty

Show this thread

And finally,

And here's

and

Here's

Claude and

ChatGPT:

Got access to Bard. Tried my fizzbuzz question: [why is the sky blue?] v0: abstained v1: hallucinated a citation. The first citation is a 404.

Show this thread

Veronica

@VeronicaSaron

BIG fan of this partnership! AI search + protection + privacy + efficiency!

Quote Tweet

Great news for @Neeva Premium members

Starting today, you can take 50% off a yearly, monthly, or family @1Password subscription!

1Password creates & stores complex, unique passwords, keeping your most important information safe and secure.

Show this thread

0:18

1.2K views

More reasons to sign up for

premium ...

Quote Tweet

Great news for @Neeva Premium members

Starting today, you can take 50% off a yearly, monthly, or family @1Password subscription!

1Password creates & stores complex, unique passwords, keeping your most important information safe and secure.

Show this thread

0:18

1.2K views

Mar 20

Predictions vs reality from

@RamaswmySridhar

on where the AI foundation models space is going ...

Quote Tweet

sridhar

@RamaswmySridhar

Mar 17

1/ And so it begins… Just as @vivek7ue and I predicted this year that @OpenAI would soon be "Closed AI"... (via @verge

) twitter.com/verge/status/1

Show this thread

Mar 17

Great conversation bt

@RamaswmySridhar

@saranormous

@eladgil

about how AI is upending search ...

Quote Tweet

Ep#8 of No-Priors.com

: @eladgil and I talk with @RamaswmySridhar, former SVP Google Ads, Commerce and Privacy and now CEO of @Neeva, about why he left Google, how AI is upending search, and the changing relationship between search and publishers/content creators.

Show this thread

0:23

1.8K views

Excited to bring

to New Zealand!

Quote Tweet

Mar 16

Hey New Zealand, @Neeva is officially available! This includes:

Access to #NeevaAI (real-time summaries backed by sources)

Zero ads

Bias-free results

Start searching at neeva.com

Show this thread

0:14

560 views

is serious about search, not ads!

Quote Tweet

What if you could control the sources that go into your AI search engine? At @Neeva, we're making that a reality. We're excited to share we added #NeevaAI:

Answer support for verified health sites, official programming sites, blogs, etc.

Availability in the News tab

Show this thread

GIF

Peter Buck

@backofthenapkin

Mar 16

How many times have you heard "trustworthy" and "search" used together? Too many times until

Neeva gets what discerning researchers want. I was using #chatgpt4 yesterday. I wanted

footnotes 2 support chat's "mansplaining" - random facts asserted w/ confidence

Quote Tweet

What if you could control the sources that go into your AI search engine? At @Neeva, we're making that a reality. We're excited to share we added #NeevaAI:

Answer support for verified health sites, official programming sites, blogs, etc.

Availability in the News tab

Show this thread

GIF

Mike Ma (Search x HCI)

@MMikeMMa

There's a HUGE gap between [LLMs for chatbots] and [LLMs in search] - -

's doing the hard work of integrating language models into search - in *reliable*, responsible, and delightful ways! Check out the series of updates they released this week! 🚀

Quote Tweet

First, at @Neeva we're passionate about generative search engines combining the best of search & AI. But it's clear generative AI systems have no notion of sources or authority. Their content is based on their reading of source material, which is often a copy of the entire Web.

Show this thread

Great BSers BS very well ... With humans, I claim there is an empirical correlation between groundedness/factuality and fluency/ability to reason ... Ques for the LLM cognoscenti -- is this also true for LLMs? If yes, what are the implications for LLMs in search ...

Claim: t5-*, flan-t5-*, ul2 and flan-ul2 are Google's greatest gifts to the ecosystem.

@YiTayML

-- any update on the timing of when the decoder variants of flan-ul2 are due for open-sourcing.

, we care to build the best search engine. For us, AI is a means to that end. We want to put you in control of your AI. Not the other way around. Starting today we are adding granular control to #NeevaAI, starting with faceted search & corpus search over news - enjoy!

Quote Tweet