The Voicebot Podcast

Project Voice gathered again in Chattanooga, Tennessee in April 2023. I had the opportunity to interview several innovative companies while on site.

• 2:16 - LilyPad AI, Genady Knizhnik - Conversational AI for English language learning
• 11:09 - Conversation Design Institute, Hans Van Dam - Conversation design evolution
• 23:11 - Speechly, Collin Borns - Speech recognition tools to combat voice chat toxicity
• 31:00 - ConverseNow, Ben Brown - Conversational AI for restaurants
• 42:51 - Calen, Irakli Beselidze - Enterprise contact center automation
• 53:42 - Dexer, Bruce Rasa - Voice AI for hands-free inspection and recording

Here is the Generative AI News (GAIN) rundown for July 20, 2023. Special segments this week include:

• Meta launches Llama 2, making it open source and free for commercial use. We talk about the implications for the LLM market.

• Did GPT-4's performance really degrade, as has been reported? We look at the research paper results, evaluate the methodology and consider a rebuttal.

• We also cover the generative AI winners and losers of the week.

My Voicebot.ai colleague Eric Schwartz joined me this week, and we went in-depth on the top news.

Links related to the stories are included below if you want to go deeper into any topics.

Top Stories of the Week

• Meta Just Became a Big Player in the LLM World, Making Llama 2 Free and Open Source

• New Research Says GPT-4 is Getting Worse. But is it True?

Generative AI Funding Fountain

• Cognaize Raises $18M to Build a Better LLM for the Finance Sector

• Preply Closes Out Series C at $120M and Doubles Down on AI

• SAP Announces Investments in Aleph Alpha, Anthropic, and Cohere

Generative AI Product Garden

• "Apple GPT" is Built on the Secret Large Language Model Ajax

• Microsoft Unveils Bing Chat Enterprise

• AI21 Labs Combats Generative AI Hallucinations With New Contextual Answers API

• Wix Text-to-Website Showcases How Copilots Streamline Complex Tasks and Can Shape Expectations

• Jasper AI Laying Off Staff 9 Months After $125M Raise

Brands in Generative AI Lands

• Associated Press Licenses its News Archive to OpenAI and will Begin Using LLMs

• Deepfake Scam Video Features British Consumer Champion Martin Lewis and Elon Musk

GAIN is recorded live and streamed via YouTube and LinkedIn on Thursdays. You can re-watch each week's discussion on Voicebot's YouTube channel. Please join us live next week on YouTube or LinkedIn. Also, please participate in the live show by commenting, and we are likely to give you a shoutout and may even show your comment on screen.

The Generative AI News (GAIN) rundown is back and was recorded on July 13, 2023. Special segments this week include:

• Anthropic's ChatGPT competitor Claude 2 landed, and it has some intriguing features. The discussion also includes a quick demo.

• Is ChatGPT's website traffic drop a reason for concern? We break down the numbers and also touch on a brand new FTC action against OpenAI.

• We also give you a new list of generative AI winners and losers of the week.

My Voicebot.ai colleague Eric Schwartz joins me today to go deep on the news. If you prefer to watch the show, you can always see the latest GAIN recording here.

Links related to the stories are included below if you want to go deeper into any topics.

Top Stories of the Week

• Anthropic's ChatGPT Alternative Claude 2 Has an Awesome New Feature and is Now Available to Everyone

• ChatGPT Web Traffic Drops. Is Generative AI's Hype Cycle Ready to Plunge?

• OpenAI Makes the GPT-4, DALL-E, Whisper, and ChatGPT Model APIs Generally Available

• FTC Investigating ChatGPT-maker OpenAI for Possible Consumer Harm

Generative AI Funding Fountain

• Voices.ai Raises $6M for Real-time AI Voice Filter

• Resemble AI Raises $8M and Launches Deepfake Voice Detector

• Generative AI Prompt Engineering Startup Vellum.ai Raises $5M

• UK Startup Prolific Raises $32M to Expand 120K-Person Network of Generative AI Model Testers

• KPMG Plans $2B Investment in AI and Cloud Services

• AI Seen Fueling Microsoft to $3T Value in Morgan Stanley Model

Generative AI Product Garden

• Shutterstock Follows Adobe and Offers Legal Protection from Generative AI Copyright Claims

• Unity Unveils Generative AI Marketplace for Video Game Developers

• X.ai to Formally Launch Business on Friday

Brands in Generative AI Lands

• Mayo Clinic Begins Testing Google Med-PaLM 2 Generative AI Model

• Gizmodo the Latest Media Publisher to Notch a Generative AI Fail

• Virgin Cruises Adopts Deepfake Jennifer Lopez as Spokesmodel

GAIN is recorded live and streamed via YouTube and LinkedIn on Thursdays. You can re-watch each week's discussion on Voicebot's YouTube channel. Please join us live next week on YouTube or LinkedIn. Also, please participate in the live show by commenting, and we are likely to give you a shoutout and may even show your comment on screen.

Please share this post with a friend, and don't forget to subscribe to the YouTube channel. Thanks!

How are ChatGPT and the rise of large language models impacting conversation designs and designers? Today, in Part 2 of our series on conversation design, we have four more experts weighing in on how the profession is changing and what designers should consider and do about it. Joining me today in a series of one-on-one interviews are:

• Maaike Coppens is vice president of product at OpenDialog AI, where she started as head of UX and conversation design in 2020. She was a product designer at XAPPMedia, labworks.io, and Smartly Ai. She was also a UX consultant at Applause. Maaike was my guest on episode 284 of the Voicebot Podcast in 2022.
• Rebecca Evanhoe is the co-author of the popular book Conversations with Things: UX Design for Chat and Voice published by Rosenfeld Media. She is a visiting assistant professor at the Pratt Institute in New York, where she teaches a graduate course in conversational UX design, was a conversation design consultant for AWS, a voice user interface designer at Mobiquity, and a writing instructor at the University of Florida. She and her co-author Diana Diebel were my guests on episode 220 in 2021.
• Kane Simms - is the founder of VUX World and a co-organizer of the Unparsed conference. VUX World is a strategic consultancy focused on conversational user experiences and is also the name of a popular podcast on practical conversational AI topics. Kane has a long history of work in technology transformation and as a writer before moving into conversation design. He was my guest in episode 178 in 2020.
• Hans Van Dam - is the CEO of the Conversation Design Institute and a co-founder and board member of the conversation design consultancy CDI Services. He is also a research fellow at RMIT University. Hans joined me on the podcast for episode 192 in 2021.

Also, the Unparsed Conference on conversation design is coming up later in July 2023 in London. If you use the code VOICEBOT, you get two tickets for the price of one. That applies to the in-person and online conference tickets. Definitely try to join online if you can't make it to London. It is a great speaker lineup. You can sign-up at https://bit.ly/unparsed2023.

Host notes - My apologies about the audio quality. We recorded on the road, and the mic setup didn't work as planned. The YouTube video audio is a little bit stronger if you want to watch this week. Also, the conversation and guest commentary is very good, so I wanted to put it out there anyway. In particular, I think you will like the discussion around LLMs and Theory of Mind, generative AI in education, and the potential erosion of generative AI startup valuations.

The Generative AI News (GAIN) rundown this week was recorded on July 6, 2023. Some special segments include:

• Mythical generative AI capabilities. From the theory of mind to general AI to emergent abilities, generative AI claims keep getting proven wrong.

• The generative AI funding fountain continues to be strong, but we discuss whether some companies are struggling to grow into their high valuations.

• We also give you a new list of generative AI winners and losers of the week.

Links related to the top stories are included below for anyone that wants to dig deeper into a topic.

Top Story of the Week

• Did ChatGPT Really Develop a Theory of Mind?

• Do Large Language Models Have Emergent Abilities?

Generative AI Funding Fountain

• Inflection AI Raises $1.3 Billion

• Runway Lands $141 Million in New Funding

• Typeface Raises $100 Million for Branded Synthetic Media Platform

• Generative AI Content Startup Scriptic Raises $5.7 Million

• Generative AI Productivity Startup Dust Raises $5.5 Million

Generative AI in the World

• German News Publisher CEO Confirms Generative AI will Replace Jobs

• 24 Top UK Universities Draft New AI Principles - Commit to Generative AI Education

GAIN is recorded live and streamed via YouTube and LinkedIn on Thursdays. You can re-watch each week's discussion on Voicebot's YouTube channel. Please join us live next week on YouTube or LinkedIn. Also, please participate in the live show by commenting, and we are likely to give you a shoutout and may even show your comment on screen.

How are ChatGPT and the rise of large language models impacting conversation designs and designers? Today, in Part 1 of our series on conversation design, we have four experts in the field weighing in on how the profession is changing and what designers should be thinking about. Joining me today in a series of one-on-one interviews are:

• Maaike Groenewege is an independent conversation designer and the creator of Convocat. She is currently the conversation design lead, prompt engineer, and NLU trainer at a U.S.-base stealth-stage startup. Groenewege has worked for 25 years as a designer, linguist, and technical writer.
• Tom Hewitson is the founder and CEO of Labworks.io, a webby-award winning conversation design studio and one of the leading voice-first game studios. Hewitson is a former journalist, content strategist at Facebook, and digital editor for Lonely Planet. He is also a lecturer at City University London and the organizer of the Unparsed Conference on conversation design.
• Peter Isaacs is the senior conversation design advocate at Voiceflow. He was previously a conversation designer at WooliesX, a division of Woolworths Group in Australia.
• Karen Kaushansky has been a conversation designer for 27 years. She is currently a Staff Conversation Designer at Google, and led the design for Google Watch.

Also, the Unparsed Conference on conversation design is coming up later in July 2023 in London. If you use the code VOICEBOT, you get two tickets for the price of one. That applies to the in-person and online conference tickets. Definitely try to join online if you can't make it to London. It is a great speaker lineup. You can sign-up at https://bit.ly/unparsed2023.

The Generative AI News (GAIN) rundown this week was recorded on June 29, 2023. Some special segments include:

• Generative AI search metrics compare wait time and word production per answer for Bing Chat, Perplexity, and Google SGE.

• The venture capital and acquisition funding fountain flowed again with a giant acquisition, financing rounds for CalypsoAI, Loora, and investment by Amazon and the UK's NHS.

• We also give you a new list of generative AI winners and losers of the week.

Links related to the top stories are included below for anyone that wants to dig deeper into a topic.

Top Story of the Week

• This One Factor Could Kill Bing Chat

• Generative AI to Reach $1 Trillion in Annual Revenue

• 47% of Technology Leaders Say AI is Top Budget Priority

Generative AI Funding Fountain

• Databricks Acquires MosaicML for $1.3 Billion

• CalypsoAI Raises $23 Million for AI Moderator and Security Solution

• Generative AI Education Startup Loora Raises $9.25 Million

• Generative AI Voice Startup Slang.ai Raises $20 Million

• Amazon Invests $100 Million in Generative AI Innovation Center

• UK Promised $27 Million to NHS for Generative AI

Generative AI Product Garden

• Oracle Introduces Generative AI for HR Productivity

• Adobe will Cover Legal Bills if You Are Sued for Copyright When Using Firefly

• Inflection AI Rolls Out In-House LLM for Enterprise Users

• Josh.ai Infuses Generative AI into Smart Home Voice Assistant

Big Brands in Generative AI Land

• AI Generates Images and Controversy for Marvel

• Wimbledon Uses IBM for Generative AI Tennis Commentary

GAIN is recorded live and streamed via YouTube and LinkedIn on Thursdays. You can re-watch each week's discussion on Voicebot's YouTube channel. Please join us live next week on YouTube or LinkedIn. Also, please participate in the live show by commenting, and we are likely to give you a shoutout and may even show your comment on screen.

Voicemod is a pioneer in audio expression. The company is best known for its real-time voice changer technologies that are popular in the gamer communities, where users can sound like a cartoon character, a movie actor, another gender, or anything you can imagine. The product's capabilities and popularity extends to other use cases as well, including use for recorded speech, sound bed generation, and more.

Jaime Bosch is CEO and co-founder of Voicemod, founded in 2014. Voicemod raised $14 million in February 2023 and is currently riding a new wave of interest brought on by the rise in generative AI. However, the company's user base growth began well before ChatGPT. Pre-pandemic, the company had 5 million users. Post-pandemic, that number was 25 million. Today Voicemod has 40 million users.

Bosch walks through the evolution of the product, including the shift from DSP to AI technology and today's hybrid solution. In addition, he shares details about the two moments of product market fit and how AI is enabling a new era of audio self-expression.

The Generative AI News (GAIN) rundown is back and was recorded on June 22, 2023. Some special segments this week include:

• The New EU AI Act and how large language models (LLM) face a big risk of non-compliance and steep fines. A new analysis breaks it down for 10 LLMs.

• The venture capital money fountain flowed with big funding rounds for Mistral, Synthesia, and ElevenLabs.

• Eric and I also bring you generative AI winners and losers of the week.

Links related to the top stories are included below for anyone that wants to dig deeper into a topic.

Top Story of the Week

• 5 Big Takeaways from the EU AI Act

• How Ready are Leading LLMs for the Draft EU AI Act (report)

Generative AI Money Fountain

• French Generative AI Startup Mistral Raises $113M

• Generative AI Speech Startup ElevenLabs Raises $19M

• Synthesia Raises $90 for Generative AI Avatar

Generative AI Product Garden

• Observe AI Dubuts Contact Center LLM

• NLX Debuts Generative AI Conversation Builder for Contact Centers

• Vimeo to Introduce Generative AI Tools for Video Production

• Meta Reveals Generative AI Synthetic Speech Tool Voicebox

Big Brands in Generative AI Land

• USA Today Publisher Gannett Taps Cohere for Generative AI Driven News Summaries

• Mercedes Adds ChatGPT to In-car Voice Assistant

GAIN is recorded live and streamed via YouTube and LinkedIn on Thursdays. You can re-watch each week's discussion on Voicebot's YouTube channel. Please join us live next week on YouTube or LinkedIn. Also, please participate in the live show by commenting, and we are likely to give you a shoutout and may even show your comment on screen.

Please share this post with a friend, and don't forget to subscribe to the YouTube channel. Thanks!

Omar Tawakol is CEO and co-founder Rembrand, a company using generative AI to automate product placement in online videos. Earlier this year, he announced an $8 million seed funding round and this month added several more strategic investors.

Rembrand represents a return to Tawakol's roots in advertising technology and his graduate work in AI. He saw there was an underutilized opportunity for product placement at scale that could be delivered cost-effectively for online influencers. AI is enabling the company to significantly expand the market.

Tawakol was my guest for Episode 61, in 2018. At the time, he was CEO of Voicera, which later rebranded to Voicea and then was acquired by Cisco 2019. Voicea was an early leader in the automated meeting transcription space and pioneered novel architectures that led to rapid increases in performance. Earlier in his career, Tawakol was CEO of Blue Kai, the leading DMP for enterprise marketers. The company was acquired by Oracle in 2016 and he then served as SVP of Oracle Data Cloud. He was formerly a board director at PlaceIQ and is a board member for Liveramp.