Ask Bob AI

Episode 103 drops today with my very dear friend, Andrew DeGood I love when my podcast guests become friends, but it might be a tiny bit more fun when my friends get to be on my podcast! Co-Founder and CEO of Ask Bob AI and IronTribe Andrew DeGood is a seasoned professional with over 20 years of experience in the finance and technology sectors. His career spans a diverse range of roles in operations, product development, client service management, and business development, culminating in the co-founding of Canopy Financial Technology Partners in 2020, which he successfully sold in 2022. Currently, Andrew is the Co-Founder and CEO of Ask Bob AI, a pioneering artificial intelligence startup that builds custom large language models for companies. With the slogan “Where Answers come faster than your morning coffee,” Ask Bob AI primarily addresses the significant time and cost organizations spend on answering questions, offering solutions that emphasize speed and cost savings. In addition to his work in AI, Andrew is passionate about mental health in the workplace and is the Co-Founder and CEO of IronTribe, a men’s community dedicated to creating connections and inspiring growth among men. Andrew’s journey includes a successful tenure at various companies, from startups to a publicly traded Fortune 500 company, where he honed his skills in strategy, team building, and storytelling to new clients and investors. His well-rounded background and dedication to cultivating amazing team cultures make him an excellent strategist and leader in his field. We cover leadership, AI, The Outlier Project, IronTribe, living with bipolar and our friendship. Don’t miss this terrific episode! #Leadership #Bipolar #MentalHealth #MensCommunity #Community #Outliers #CEO

While Microsoft made the right choice with OpenAI, they have seen continued struggles with co-pilot which continues to take blows with the latest around security concerns with Recall. In essence Recall records your desktop screen and then allows you to “recall” anything you’ve ever done including web searches, documents, etc. And it continues to show it is super easy to hack. Simply put I would not use Recall and I definitely wouldn’t deploy it within my company. Just something to be aware of as you proceed with your company’s AI plan. Happy Saturday all. #AI #SecurityConcerns #Microsoft #OpenAI #CoPilot #Recall #TechNews #CyberSecurity #BusinessStrategy #AIPlanning #TechAwareness #SaturdayThoughts

Some thoughts on Colorado Bill 205 (Consumer Protections for Artificial Intelligence) which passed last month and is the first comprehensive AI legislation in the country. It is likely to set forth a framework that other states follow so no matter where you are it is a good standard to at least design around until we see what comes from other states. One major thing though is this, just because you are not an AI developer does not mean you are at risk! Clients of AI SAAS will likely also carry liability if they use what Colorado calls a high risk AI system. Which include the following and possibly more: •Employment (hiring, firing and promotions) •Education (admissions and scholarships) •Lending (loan approvals, credit scoring) •Financial Services (underwriting and pricing) •Healthcare (diagnosis, treatment, recommendations) And will also be required to disclose any use of AI to their clients. Two things mortgage companies should start doing asap is first, ensure you are disclosing any use of AI to your customers including marketing content generated or any systems within your process. Secondly prepare a plan for how you will test the systems you use that are considered high risk, for potential discrimination. You shouldn’t leave this up to your vendor. Now as a disclaimer I’m obviously not a lawyer. But I know a few 😉 Wendy Lee, Hailey Rice, Jeremy Potter and John Levonick I’d love to have you chime in here. Happy Friday all. #AI #ArtificialIntelligence #ColoradoBill205 #ConsumerProtection #AIGovernance #HighRiskAI #AIRegulation #AICompliance #AILiability #Fintech #HealthcareAI #EducationAI #EmploymentAI #MortgageIndustry #DiscloseAI #AIEthics #AIinFinance #AIinHealthcare #AIinEducation #AIinEmployment #TechLaw #AITesting #AIDiscrimination #FridayThoughts

Best tweet of the week and shows how dependent a smallish portion of the world is on GPT and shows OpenAI’s strength in the market. See my other post from today, it’s basically like the power lines to your house went down and nothing works. Did you notice? Did it impact your day? Or are you just now hearing about it? #OpenAI #GPT #AI #TechDependence #MarketLeader #Innovation #TechPower #FutureOfAI #AIImpact #TechReliance

This week it’s all about Microsoft as their build conference is going on. It’s clearly announcement season. I wanted to quickly share this demo on onscreen co-pilot which is powered by GPT-4o. While the voice needs some work and don’t worry it will improve quickly, this will be you speed up your interactions with your own personal assistant. Take a watch while the AI assistant helps this dad play Minecraft. The use cases are limitless. More to come from Microsoft this week! Happy Tuesday all. #MicrosoftBuild #GPT4 #AIAssistant #TechInnovation #Microsoft #ArtificialIntelligence #CoPilot #VoiceTechnology #ProductivityBoost #TechTrends #Innovation #PersonalAssistant #TechDemo #FutureOfTech

Interesting item that I’ve yet to see anyone bring up. Why did OpenAI launch their desktop app on a Mac before a Microsoft powered device? Does OpenAI have eyes for a new bigger partner? Or is Microsoft starting to align with Apple? 🤔 #OpenAI #AI #ArtificialIntelligence #Mac #Microsoft #Apple #TechTrends #BusinessStrategy #Partnerships #TechNews #Innovation #FutureOfTech #AIIndustry #Technology

Quick hot take on the barage of resignations at OpenAI the past two days. One possible theory is OpenAI has achieved AGI but are suppressing it for the time being. I believe most of the people resigning would see no reason to remain if AGI or ASI has been achieved. 🤔 But there is no shortage of theories at this point. #openai #resignations #AGI #ASI

Google did such an amazing job yesterday that I feel badly for them, it felt like OpenAI stole most of their thunder the day prior. We are fascinated with the emerging new tech that feels magical! And I think OpenAI just stole some of the magic by going first this week. So with that here are eight of the impressive but not so magical releases yesterday. 👨💻 Project Astra: This is Google's new AI Assistant which is powered by Gemini and has both vision and voice. The demos were super impressive, albeit we didn't see a demo of the emotion components that OpenAI showed which I think is the bigger leap in understanding. 👩💻 Veo: Google's new text to video creation tool, this is similar to Sora from OpenAI although it did appear Veo creates videos longer than Sora but we won't really know till both of them release to the public. 👨💻 AI native in search: This was coming and had to come. Google has known it needed to quickly update search and they did. This is Google's bread and butter and know they spend a ton of energy on this. 👩💻 Imagen 3: This an upgrade to their text to image generation, not able to be tested yet. 👨💻 Google Workspace: This is the competitor to Microsoft Co-Pilot and was a super impressive demo on how you can search all of your emails, documents, etc. It will be $20 a month and will include Gemini Pro 1.5. If you are a heavy Google user it is worth it. It has some pretty cool automations built in. 👩💻 Gemini 1.5: While this model has been talked about for a while they upped the context window to 2 million tokens which is approximately 1.5 million words. This is a huge leap in the amount of information the model can process in one setting. For context all of the Harry Potter books combined only have a little over 1 million words. And GPT-4 has a context window of 32K tokens on chat and 128k tokens on the API. More importantly Google's CEO made the obvious statement that everyone is chasing an infinite context window. 👩💻 Google Agents: Similar to the GPT store we will see Google Agents coming along with a Gemini App directly on your phone. Allowing developers to build on top of the established Gemini Assistant. 👨💻 : The next Pixel phone will have a nano multimodal LLM built in. This will make AI engagements much quicker as many tasks will be native. Expect Apple to do the same thing in September. What was your favorite thing Google released yesterday? Or do you feel I missed something? Let me know down in the comments 👇 #TechNews #ArtificialIntelligence #AI #Google #OpenAI #ProjectAstra #Veo #AIAssistants #AIinSearch #Imagen3 #GoogleWorkspace #Gemini #GoogleAgents #PixelPhone #EmergingTech #Innovation #TechReleases

I am still trying to wrap my head around all of OpenAI's updates yesterday. The volumes of applications of the voice and vision are almost overwhelming. It also made me realize OpenAI's potential strategy to make it sticky with customers is the memory feature roll out a few weeks back that I did a post on in combination with voice. The ability to detect emotion and express emotion, along with real time interaction is not something I expected we would see this soon. The next step is GPT will be able to place phone calls with for you. Which brings us back to the memory thing. As you teach GPT more and more about you it becomes a more able bodied assistant, just like a real life assistant. Let's assume it now knows your schedule (calendar integrations coming soon, I'd bank on it) your name, your dog, your vet, and some other details. Now you can just tell ChatGPT to book you a vet appointment for your dog. ChatGPT will be able to call your vet directly and book the appointment and then place it into your calendar and notify you. Now imagine that scenario but for all the calls that are a waste of your time. ChatGPT doesn't mind sitting on hold for 20 minutes. 😉 PS and now the vet can have an AI voice assistant on the other side that can receive the calls and book the appointments. Next Up: No rest on AI this week. Today we have Google I/O event at 1pm EST, I will be back this afternoon for a quick breakdown of that event as well so you don't have to watch the whole thing. 📸 : The video below is a demo from OpenAI that was not part of the spring update. It shows two GPTs talking to each other. 🤯 🤯 🤯 One of the GPTs is using vision capabilities and the other is directed them with the camera to explore the room. Happy Tuesday all. #OpenAI #AIUpdates #VoiceAI #VisionAI #RealTimeAI #ArtificialIntelligence #AIAssistant #TechTrends #AIInnovation #GPT4o #DigitalTransformation

OpenAI just finished their spring update but before I get into the updates themselves I'd like to note the funniest thing about them. The used a Mac for all the demos, seems kind of hilarious given Microsoft is their biggest backer and makes me happy I own a Macbook. 😂 There were 4 core components to the rollout, a new free model GPT-4o, a new desktop version, enhanced voice capabilities and enhanced vision capabilities. GPT-4o : This model brings GPT-4 level quality to free users, a mission statement that OpenAI has continued to stick with as their belief these tools should be available to the general public. It is much faster than GPT-4 and major enhancements on the API which is now 2x faster, 50% cheaper and 5x on rate limits. There is still many reasons to keep your paid account, most of all if you are a heavy user of the application like I am. Speech : This blew me away in the demo. The voice function is so fast now, you can easily have a nature language conversation with GPT and the voice now detects emotion in your voice and expresses emotion in it's own voice. This is going to have major ramifications in its applications within therapy and as a companion (I know we are all scared of the movie Her, fair point). But the demo of this function was super impressive. Also they demoed in real time the ability for real time translation in voice. They carried on a conversation between Italian and English and it nailed it out of the park. Vision : You can use your phone camera to interact with GPT now and you can use the desktop app to allow GPT to view your computer desktop to help you with work in real time. The voice capabilities are also on the desktop version. This makes GPT-4o an amazing application for tutoring and the perfect co-pilot for your day to day work tasks. Desktop : Until now you had to open your browser to access GPT-4 and it has obvious limitations as browser only based. With the desktop app you can now utilize GPT-4o with both voice and vision directly on your desktop, and as mentioned above you can see the huge amount of applications for it. These features will begin rolling out in the coming weeks, so like memory it could be a hot minute till you see these come out but as they do I will make sure to let you know. Now onto the Google I/O event tomorrow. It is a crazy, insane week in AI this week. Hold on to your hats for more to come tomorrow afternoon. #OpenAIUpdate #GPT4o #AIForEveryone #VoiceTechnology #VisionAI #DesktopAI #TechInnovation #AICompanion #RealTimeTranslation #ArtificialIntelligence